Publications

State-of-the-art novel view synthesis methods achieve impressive results for
multi-view captures of static 3D scenes. However, the reconstructed scenes
still lack "liveliness," a key component for creating engaging 3D experiences.
Recently, novel video diffusion models generate realistic videos with complex
motion and enable animations of 2D images, however they cannot naively be used
to animate 3D scenes as they lack multi-view consistency. To breathe life into
the static world, we propose Gaussians2Life, a method for animating parts of
high-quality 3D scenes in a Gaussian Splatting representation. Our key idea is
to leverage powerful video diffusion models as the generative component of our
model and to combine these with a robust technique to lift 2D videos into
meaningful 3D motion. We find that, in contrast to prior work, this enables
realistic animations of complex, pre-existing 3D scenes and further enables the
animation of a large variety of object classes, while related work is mostly
focused on prior-based character animation, or single 3D objects. Our model
enables the creation of consistent, immersive 3D experiences for arbitrary
scenes.

We introduce PersonaHOI, a training- and tuning-free framework that fuses a
general StableDiffusion model with a personalized face diffusion (PFD) model to
generate identity-consistent human-object interaction (HOI) images. While
existing PFD models have advanced significantly, they often overemphasize
facial features at the expense of full-body coherence, PersonaHOI introduces an
additional StableDiffusion (SD) branch guided by HOI-oriented text inputs. By
incorporating cross-attention constraints in the PFD branch and spatial merging
at both latent and residual levels, PersonaHOI preserves personalized facial
details while ensuring interactive non-facial regions. Experiments, validated
by a novel interaction alignment metric, demonstrate the superior realism and
scalability of PersonaHOI, establishing a new standard for practical
personalized face with HOI generation. Our code will be available at
github.com/JoyHuYY1412/PersonaHOI

We introduce PersonaHOI, a training- and tuning-free framework that fuses a
general StableDiffusion model with a personalized face diffusion (PFD) model to
generate identity-consistent human-object interaction (HOI) images. While
existing PFD models have advanced significantly, they often overemphasize
facial features at the expense of full-body coherence, PersonaHOI introduces an
additional StableDiffusion (SD) branch guided by HOI-oriented text inputs. By
incorporating cross-attention constraints in the PFD branch and spatial merging
at both latent and residual levels, PersonaHOI preserves personalized facial
details while ensuring interactive non-facial regions. Experiments, validated
by a novel interaction alignment metric, demonstrate the superior realism and
scalability of PersonaHOI, establishing a new standard for practical
personalized face with HOI generation. Our code will be available at
github.com/JoyHuYY1412/PersonaHOI

With the rise of generative AI, synthesizing figures from text captions
becomes a compelling application. However, achieving high geometric precision
and editability requires representing figures as graphics programs in languages
like TikZ, and aligned training data (i.e., graphics programs with captions)
remains scarce. Meanwhile, large amounts of unaligned graphics programs and
captioned raster images are more readily available. We reconcile these
disparate data sources by presenting TikZero, which decouples graphics program
generation from text understanding by using image representations as an
intermediary bridge. It enables independent training on graphics programs and
captioned images and allows for zero-shot text-guided graphics program
synthesis during inference. We show that our method substantially outperforms
baselines that can only operate with caption-aligned graphics programs.
Furthermore, when leveraging caption-aligned graphics programs as a
complementary training signal, TikZero matches or exceeds the performance of
much larger models, including commercial systems like GPT-4o. Our code,
datasets, and select models are publicly available.

Flow-based latent generative models such as Stable Diffusion 3 are able to
generate images with remarkable quality, even enabling photorealistic
text-to-image generation. Their impressive performance suggests that these
models should also constitute powerful priors for inverse imaging problems, but
that approach has not yet led to comparable fidelity. There are several key
obstacles: (i) the encoding into a lower-dimensional latent space makes the
underlying (forward) mapping non-linear; (ii) the data likelihood term is
usually intractable; and (iii) learned generative models struggle to recover
rare, atypical data modes during inference. We present FLAIR, a novel training
free variational framework that leverages flow-based generative models as a
prior for inverse problems. To that end, we introduce a variational objective
for flow matching that is agnostic to the type of degradation, and combine it
with deterministic trajectory adjustments to recover atypical modes. To enforce
exact consistency with the observed data, we decouple the optimization of the
data fidelity and regularization terms. Moreover, we introduce a time-dependent
calibration scheme in which the strength of the regularization is modulated
according to off-line accuracy estimates. Results on standard imaging
benchmarks demonstrate that FLAIR consistently outperforms existing diffusion-
and flow-based methods in terms of reconstruction quality and sample diversity.

Neural networks are widely adopted to solve complex and challenging tasks.
Especially in high-stakes decision-making, understanding their reasoning
process is crucial, yet proves challenging for modern deep networks. Feature
visualization (FV) is a powerful tool to decode what information neurons are
responding to and hence to better understand the reasoning behind such
networks. In particular, in FV we generate human-understandable images that
reflect the information detected by neurons of interest. However, current
methods often yield unrecognizable visualizations, exhibiting repetitive
patterns and visual artifacts that are hard to understand for a human. To
address these problems, we propose to guide FV through statistics of real image
features combined with measures of relevant network flow to generate
prototypical images. Our approach yields human-understandable visualizations
that both qualitatively and quantitatively improve over state-of-the-art FVs
across various architectures. As such, it can be used to decode which
information the network uses, complementing mechanistic circuits that identify
where it is encoded. Code is available at: github.com/adagorgun/VITAL

Deep learning has become an essential part of computer vision, with deep
neural networks (DNNs) excelling in predictive performance. However, they often
fall short in other critical quality dimensions, such as robustness,
calibration, or fairness. While existing studies have focused on a subset of
these quality dimensions, none have explored a more general form of
"well-behavedness" of DNNs. With this work, we address this gap by
simultaneously studying nine different quality dimensions for image
classification. Through a large-scale study, we provide a bird's-eye view by
analyzing 326 backbone models and how different training paradigms and model
architectures affect the quality dimensions. We reveal various new insights
such that (i) vision-language models exhibit high fairness on ImageNet-1k
classification and strong robustness against domain changes; (ii)
self-supervised learning is an effective training paradigm to improve almost
all considered quality dimensions; and (iii) the training dataset size is a
major driver for most of the quality dimensions. We conclude our study by
introducing the QUBA score (Quality Understanding Beyond Accuracy), a novel
metric that ranks models across multiple dimensions of quality, enabling
tailored recommendations based on specific user needs.

Deep neural networks (DNNs) have proven to be successful in various computer
vision applications such that models even infer in safety-critical situations.
Therefore, vision models have to behave in a robust way to disturbances such as
noise or blur. While seminal benchmarks exist to evaluate model robustness to
diverse corruptions, blur is often approximated in an overly simplistic way to
model defocus, while ignoring the different blur kernel shapes that result from
optical systems. To study model robustness against realistic optical blur
effects, this paper proposes two datasets of blur corruptions, which we denote
OpticsBench and LensCorruptions. OpticsBench examines primary aberrations such
as coma, defocus, and astigmatism, i.e. aberrations that can be represented by
varying a single parameter of Zernike polynomials. To go beyond the principled
but synthetic setting of primary aberrations, LensCorruptions samples linear
combinations in the vector space spanned by Zernike polynomials, corresponding
to 100 real lenses. Evaluations for image classification and object detection
on ImageNet and MSCOCO show that for a variety of different pre-trained models,
the performance on OpticsBench and LensCorruptions varies significantly,
indicating the need to consider realistic image corruptions to evaluate a
model's robustness against blur.

With the rise of neural networks, especially in high-stakes applications,
these networks need two properties (i) robustness and (ii) interpretability to
ensure their safety. Recent advances in classifiers with 3D volumetric object
representations have demonstrated a greatly enhanced robustness in
out-of-distribution data. However, these 3D-aware classifiers have not been
studied from the perspective of interpretability. We introduce CAVE - Concept
Aware Volumes for Explanations - a new direction that unifies interpretability
and robustness in image classification. We design an inherently-interpretable
and robust classifier by extending existing 3D-aware classifiers with concepts
extracted from their volumetric representations for classification. In an array
of quantitative metrics for interpretability, we compare against different
concept-based approaches across the explainable AI literature and show that
CAVE discovers well-grounded concepts that are used consistently across images,
while achieving superior robustness.

Monocular 3D estimation is crucial for visual perception. However, current
methods fall short by relying on oversimplified assumptions, such as pinhole
camera models or rectified images. These limitations severely restrict their
general applicability, causing poor performance in real-world scenarios with
fisheye or panoramic images and resulting in substantial context loss. To
address this, we present UniK3D, the first generalizable method for monocular
3D estimation able to model any camera. Our method introduces a spherical 3D
representation which allows for better disentanglement of camera and scene
geometry and enables accurate metric 3D reconstruction for unconstrained camera
models. Our camera component features a novel, model-independent representation
of the pencil of rays, achieved through a learned superposition of spherical
harmonics. We also introduce an angular loss, which, together with the camera
module design, prevents the contraction of the 3D outputs for wide-view
cameras. A comprehensive zero-shot evaluation on 13 diverse datasets
demonstrates the state-of-the-art performance of UniK3D across 3D, depth, and
camera metrics, with substantial gains in challenging large-field-of-view and
panoramic settings, while maintaining top accuracy in conventional pinhole
small-field-of-view domains. Code and models are available at
github.com/lpiccinelli-eth/unik3d .

Accurate monocular metric depth estimation (MMDE) is crucial to solving
downstream tasks in 3D perception and modeling. However, the remarkable
accuracy of recent MMDE methods is confined to their training domains. These
methods fail to generalize to unseen domains even in the presence of moderate
domain gaps, which hinders their practical applicability. We propose a new
model, UniDepthV2, capable of reconstructing metric 3D scenes from solely
single images across domains. Departing from the existing MMDE paradigm,
UniDepthV2 directly predicts metric 3D points from the input image at inference
time without any additional information, striving for a universal and flexible
MMDE solution. In particular, UniDepthV2 implements a self-promptable camera
module predicting a dense camera representation to condition depth features.
Our model exploits a pseudo-spherical output representation, which disentangles
the camera and depth representations. In addition, we propose a geometric
invariance loss that promotes the invariance of camera-prompted depth features.
UniDepthV2 improves its predecessor UniDepth model via a new edge-guided loss
which enhances the localization and sharpness of edges in the metric depth
outputs, a revisited, simplified and more efficient architectural design, and
an additional uncertainty-level output which enables downstream tasks requiring
confidence. Thorough evaluations on ten depth datasets in a zero-shot regime
consistently demonstrate the superior performance and generalization of
UniDepthV2. Code and models are available at
github.com/lpiccinelli-eth/UniDepth

Climate change is one of the most pressing challenges of the 21st century,
sparking widespread discourse across social media platforms. Activists,
policymakers, and researchers seek to understand public sentiment and
narratives while access to social media data has become increasingly restricted
in the post-API era. In this study, we analyze a dataset of climate
change-related tweets from X (formerly Twitter) shared in 2019, containing 730k
tweets along with the shared images. Our approach integrates statistical
analysis, image classification, object detection, and sentiment analysis to
explore visual narratives in climate discourse. Additionally, we introduce a
graphical user interface (GUI) to facilitate interactive data exploration. Our
findings reveal key themes in climate communication, highlight sentiment
divergence between images and text, and underscore the strengths and
limitations of foundation models in analyzing social media imagery. By
releasing our code and tools, we aim to support future research on the
intersection of climate change, social media, and computer vision.

Concept Bottleneck Models (CBMs) enhance the interpretability of neural
networks by basing predictions on human-understandable concepts. However,
current CBMs typically rely on concept sets extracted from large language
models or extensive image corpora, limiting their effectiveness in data-sparse
scenarios. We propose Data-efficient CBMs (DCBMs), which reduce the need for
large sample sizes during concept generation while preserving interpretability.
DCBMs define concepts as image regions detected by segmentation or detection
foundation models, allowing each image to generate multiple concepts across
different granularities. This removes reliance on textual descriptions and
large-scale pre-training, making DCBMs applicable for fine-grained
classification and out-of-distribution tasks. Attribution analysis using
Grad-CAM demonstrates that DCBMs deliver visual concepts that can be localized
in test images. By leveraging dataset-specific concepts instead of predefined
ones, DCBMs enhance adaptability to new domains.

Open-vocabulary semantic segmentation models associate vision and text to
label pixels from an undefined set of classes using textual queries, providing
versatile performance on novel datasets. However, large shifts between training
and test domains degrade their performance, requiring fine-tuning for effective
real-world applications. We introduce Semantic Library Adaptation (SemLA), a
novel framework for training-free, test-time domain adaptation. SemLA leverages
a library of LoRA-based adapters indexed with CLIP embeddings, dynamically
merging the most relevant adapters based on proximity to the target domain in
the embedding space. This approach constructs an ad-hoc model tailored to each
specific input without additional training. Our method scales efficiently,
enhances explainability by tracking adapter contributions, and inherently
protects data privacy, making it ideal for sensitive applications.
Comprehensive experiments on a 20-domain benchmark built over 10 standard
datasets demonstrate SemLA's superior adaptability and performance across
diverse settings, establishing a new standard in domain adaptation for
open-vocabulary semantic segmentation.

Post-hoc explanation methods for black-box models often struggle with
faithfulness and human interpretability due to the lack of explainability in
current neural models. Meanwhile, B-cos networks have been introduced to
improve model explainability through architectural and computational
adaptations, but their application has so far been limited to computer vision
models and their associated training pipelines. In this work, we introduce
B-cos LMs, i.e., B-cos networks empowered for NLP tasks. Our approach directly
transforms pre-trained language models into B-cos LMs by combining B-cos
conversion and task fine-tuning, improving efficiency compared to previous
B-cos methods. Our automatic and human evaluation results demonstrate that
B-cos LMs produce more faithful and human interpretable explanations than post
hoc methods, while maintaining task performance comparable to conventional
fine-tuning. Our in-depth analysis explores how B-cos LMs differ from
conventionally fine-tuned models in their learning processes and explanation
patterns. Finally, we provide practical guidelines for effectively building
B-cos LMs based on our findings. Our code is available at
anonymous.4open.science/r/bcos_lm.

We introduce Spatial Reasoning Models (SRMs), a framework to perform
reasoning over sets of continuous variables via denoising generative models.
SRMs infer continuous representations on a set of unobserved variables, given
observations on observed variables. Current generative models on spatial
domains, such as diffusion and flow matching models, often collapse to
hallucination in case of complex distributions. To measure this, we introduce a
set of benchmark tasks that test the quality of complex reasoning in generative
models and can quantify hallucination. The SRM framework allows to report key
findings about importance of sequentialization in generation, the associated
order, as well as the sampling strategies during training. It demonstrates, for
the first time, that order of generation can successfully be predicted by the
denoising network itself. Using these findings, we can increase the accuracy of
specific reasoning tasks from 1% to >50%.

Recent advances in multi-modal generative models have enabled significant
progress in instruction-based image editing. However, while these models
produce visually plausible outputs, their capacity for knowledge-based
reasoning editing tasks remains under-explored. In this paper, we introduce
KRIS-Bench (Knowledge-based Reasoning in Image-editing Systems Benchmark), a
diagnostic benchmark designed to assess models through a cognitively informed
lens. Drawing from educational theory, KRIS-Bench categorizes editing tasks
across three foundational knowledge types: Factual, Conceptual, and Procedural.
Based on this taxonomy, we design 22 representative tasks spanning 7 reasoning
dimensions and release 1,267 high-quality annotated editing instances. To
support fine-grained evaluation, we propose a comprehensive protocol that
incorporates a novel Knowledge Plausibility metric, enhanced by knowledge hints
and calibrated through human studies. Empirical results on 10 state-of-the-art
models reveal significant gaps in reasoning performance, highlighting the need
for knowledge-centric benchmarks to advance the development of intelligent
image editing systems.

B-cos Networks have been shown to be effective for obtaining highly human
interpretable explanations of model decisions by architecturally enforcing
stronger alignment between inputs and weight. B-cos variants of convolutional
networks (CNNs) and vision transformers (ViTs), which primarily replace linear
layers with B-cos transformations, perform competitively to their respective
standard variants while also yielding explanations that are faithful by design.
However, it has so far been necessary to train these models from scratch, which
is increasingly infeasible in the era of large, pre-trained foundation models.
In this work, inspired by the architectural similarities in standard DNNs and
B-cos networks, we propose 'B-cosification', a novel approach to transform
existing pre-trained models to become inherently interpretable. We perform a
thorough study of design choices to perform this conversion, both for
convolutional neural networks and vision transformers. We find that
B-cosification can yield models that are on par with B-cos models trained from
scratch in terms of interpretability, while often outperforming them in terms
of classification performance at a fraction of the training cost. Subsequently,
we apply B-cosification to a pretrained CLIP model, and show that, even with
limited data and compute cost, we obtain a B-cosified version that is highly
interpretable and competitive on zero shot performance across a variety of
datasets. We release our code and pre-trained model weights at
github.com/shrebox/B-cosification.

Hand action recognition is essential. Communication, human-robot
interactions, and gesture control are dependent on it. Skeleton-based action
recognition traditionally includes hands, which belong to the classes which
remain challenging to correctly recognize to date. We propose a method
specifically designed for hand action recognition which uses relative angular
embeddings and local Spherical Harmonics to create novel hand representations.
The use of Spherical Harmonics creates rotation-invariant representations which
make hand action recognition even more robust against inter-subject differences
and viewpoint changes. We conduct extensive experiments on the hand joints in
the First-Person Hand Action Benchmark with RGB-D Videos and 3D Hand Pose
Annotations, and on the NTU RGB+D 120 dataset, demonstrating the benefit of
using Local Spherical Harmonics Representations. Our code is available at
github.com/KathPra/LSHR_LSHT.

Image restoration networks are usually comprised of an encoder and a decoder,
responsible for aggregating image content from noisy, distorted data and to
restore clean, undistorted images, respectively. Data aggregation as well as
high-resolution image generation both usually come at the risk of involving
aliases, i.e.~standard architectures put their ability to reconstruct the model
input in jeopardy to reach high PSNR values on validation data. The price to be
paid is low model robustness. In this work, we show that simply providing
alias-free paths in state-of-the-art reconstruction transformers supports
improved model robustness at low costs on the restoration performance. We do so
by proposing BOA-Restormer, a transformer-based image restoration model that
executes downsampling and upsampling operations partly in the frequency domain
to ensure alias-free paths along the entire model while potentially preserving
all relevant high-frequency information.

We introduce SLayR, Scene Layout Generation with Rectified flow.
State-of-the-art text-to-image models achieve impressive results. However, they
generate images end-to-end, exposing no fine-grained control over the process.
SLayR presents a novel transformer-based rectified flow model for layout
generation over a token space that can be decoded into bounding boxes and
corresponding labels, which can then be transformed into images using existing
models. We show that established metrics for generated images are inconclusive
for evaluating their underlying scene layout, and introduce a new benchmark
suite, including a carefully designed repeatable human-evaluation procedure
that assesses the plausibility and variety of generated layouts. In contrast to
previous works, which perform well in either high variety or plausibility, we
show that our approach performs well on both of these axes at the same time. It
is also at least 5x times smaller in the number of parameters and 37% faster
than the baselines. Our complete text-to-image pipeline demonstrates the added
benefits of an interpretable and editable intermediate representation.

Low-dimensional embeddings (LDEs) of high-dimensional data are ubiquitous in
science and engineering. They allow us to quickly understand the main
properties of the data, identify outliers and processing errors, and inform the
next steps of data analysis. As such, LDEs have to be faithful to the original
high-dimensional data, i.e., they should represent the relationships that are
encoded in the data, both at a local as well as global scale. The current
generation of LDE approaches focus on reconstructing local distances between
any pair of samples correctly, often out-performing traditional approaches
aiming at all distances. For these approaches, global relationships are,
however, usually strongly distorted, often argued to be an inherent trade-off
between local and global structure learning for embeddings. We suggest a new
perspective on LDE learning, reconstructing angles between data points. We show
that this approach, Mercat, yields good reconstruction across a diverse set of
experiments and metrics, and preserve structures well across all scales.
Compared to existing work, our approach also has a simple formulation,
facilitating future theoretical analysis and algorithmic improvements.

Vision language models (VLMs) have drastically changed the computer vision
model landscape in only a few years, opening an exciting array of new
applications from zero-shot image classification, over to image captioning, and
visual question answering. Unlike pure vision models, they offer an intuitive
way to access visual content through language prompting. The wide applicability
of such models encourages us to ask whether they also align with human vision -
specifically, how far they adopt human-induced visual biases through multimodal
fusion, or whether they simply inherit biases from pure vision models. One
important visual bias is the texture vs. shape bias, or the dominance of local
over global information. In this paper, we study this bias in a wide range of
popular VLMs. Interestingly, we find that VLMs are often more shape-biased than
their vision encoders, indicating that visual biases are modulated to some
extent through text in multimodal models. If text does indeed influence visual
biases, this suggests that we may be able to steer visual biases not just
through visual input but also through language: a hypothesis that we confirm
through extensive experiments. For instance, we are able to steer shape bias
from as low as 49% to as high as 72% through prompting alone. For now, the
strong human bias towards shape (96%) remains out of reach for all tested VLMs.

With the rapid growth of the volume of research fields like computer vision
and computer graphics, researchers require effective and user-friendly
rendering tools to visualize results. While advanced tools like Blender offer
powerful capabilities, they also require a significant effort to master. This
technical report introduces Blendify, a lightweight Python-based framework that
seamlessly integrates with Blender, providing a high-level API for scene
creation and rendering. Blendify reduces the complexity of working with
Blender's native API by automating object creation, handling the colors and
material linking, and implementing features such as shadow-catcher objects
while maintaining support for high-quality ray-tracing rendering output. With a
focus on usability Blendify enables efficient and flexible rendering workflow
for rendering in common computer vision and computer graphics use cases. The
code is available at github.com/ptrvilya/blendify

The estimation of 3D human poses from images has progressed tremendously over
the last few years as measured on standard benchmarks. However, performance in
the open world remains underexplored, as current benchmarks cannot capture its
full extent. Especially in safety-critical systems, it is crucial that 3D pose
estimators are audited before deployment, and their sensitivity towards single
factors or attributes occurring in the operational domain is thoroughly
examined. Nevertheless, we currently lack a benchmark that would enable such
fine-grained analysis. We thus present STAGE, a GenAI data toolkit for auditing
3D human pose estimators. We enable a text-to-image model to control the 3D
human body pose in the generated image. This allows us to create customized
annotated data covering a wide range of open-world attributes. We leverage
STAGE and generate a series of benchmarks to audit the sensitivity of popular
pose estimators towards attributes such as gender, ethnicity, age, clothing,
location, and weather. Our results show that the presence of such naturally
occurring attributes can cause severe degradation in the performance of pose
estimators and leads us to question if they are ready for open-world
deployment.

Deep learning is increasingly relevant in our daily lives, as it simplifies tedious tasks and enhances quality of life across various domains such as entertainment, learning, automatic assistance, and autonomous driving. However, the demand for more data to train models for emerging tasks is increasing dramatically. Deep learning models heavily depend on the quality and quantity of data, necessitating high-quality labeled datasets. Yet, each task requires different types of annotations for training and evaluation, posing challenges in obtaining comprehensive supervision. The acquisition of annotations is not only resource-intensive in terms of time and cost but also introduces biases, such as granularity in classification, where distinctions like specific breeds versus generic categories may arise. Furthermore, the dynamic nature of the world causes the challenge that previously annotated data becomes potentially irrelevant, and new categories and rare occurrences continually emerge, making it impossible to label every aspect of the world.
Therefore, this thesis aims to explore various supervision scenarios to mitigate the need for full supervision and reduce data acquisition costs. Specifically, we investigate learning without labels, referred to as self-supervised and unsupervised methods, to better understand video and image representations. To learn from data without labels, we leverage injected priors such as motion speed, direction, action order in videos, or semantic information granularity to obtain powerful data representations. Further, we study scenarios involving reduced supervision levels. To reduce annotation costs, first, we propose to omit precise annotations for one modality in multimodal learning, namely in text-video and image-video settings, and transfer available knowledge to large copora of video data. Second, we study semi-supervised learning scenarios, where only a subset of annotated data alongside unlabeled data is available, and propose to revisit regularization constraints and improve generalization to unlabeled data. Additionally, we address scenarios where parts of available data is inherently limited due to privacy and security reasons or naturally rare events, which not only restrict annotations but also limit the overall data volume. For these scenarios, we propose methods that carefully balance between previously obtained knowledge and incoming limited data by introducing a calibration method or combining a space reservation technique with orthogonality constraints. Finally, we explore multimodal and unimodal open-world scenarios where the model is asked to generalize beyond the given set of object or action classes. Specifically, we propose a new challenging setting on multimodal egocentric videos and propose an adaptation method for vision-language models to generalize on egocentric domain. Moreover, we study unimodal image recognition in an open-set setting and propose to disentangle open-set detection and image classification tasks that effectively improve generalization in different settings.
In summary, this thesis investigates challenges arising when full supervision for training models is not available. We develop methods to understand learning dynamics and the role of biases in data, while also proposing novel setups to advance training with less supervision.

Reasoning over sports videos for question answering is an important task with
numerous applications, such as player training and information retrieval.
However, this task has not been explored due to the lack of relevant datasets
and the challenging nature it presents. Most datasets for video question
answering (VideoQA) focus mainly on general and coarse-grained understanding of
daily-life videos, which is not applicable to sports scenarios requiring
professional action understanding and fine-grained motion analysis. In this
paper, we introduce the first dataset, named Sports-QA, specifically designed
for the sports VideoQA task. The Sports-QA dataset includes various types of
questions, such as descriptions, chronologies, causalities, and counterfactual
conditions, covering multiple sports. Furthermore, to address the
characteristics of the sports VideoQA task, we propose a new Auto-Focus
Transformer (AFT) capable of automatically focusing on particular scales of
temporal information for question answering. We conduct extensive experiments
on Sports-QA, including baseline studies and the evaluation of different
methods. The results demonstrate that our AFT achieves state-of-the-art
performance.

Autoregressive (AR) models have achieved remarkable success in natural
language and image generation, but their application to 3D shape modeling
remains largely unexplored. Unlike diffusion models, AR models enable more
efficient and controllable generation with faster inference times, making them
especially suitable for data-intensive domains. Traditional 3D generative
models using AR approaches often rely on ``next-token" predictions at the voxel
or point level. While effective for certain applications, these methods can be
restrictive and computationally expensive when dealing with large-scale 3D
data. To tackle these challenges, we introduce 3D-WAG, an AR model for 3D
implicit distance fields that can perform unconditional shape generation,
class-conditioned and also text-conditioned shape generation. Our key idea is
to encode shapes as multi-scale wavelet token maps and use a Transformer to
predict the ``next higher-resolution token map" in an autoregressive manner. By
redefining 3D AR generation task as ``next-scale" prediction, we reduce the
computational cost of generation compared to traditional ``next-token"
prediction models, while preserving essential geometric details of 3D shapes in
a more structured and hierarchical manner. We evaluate 3D-WAG to showcase its
benefit by quantitative and qualitative comparisons with state-of-the-art
methods on widely used benchmarks. Our results show 3D-WAG achieves superior
performance in key metrics like Coverage and MMD, generating high-fidelity 3D
shapes that closely match the real data distribution.

While being very successful in solving many downstream tasks, the application
of deep neural networks is limited in real-life scenarios because of their
susceptibility to domain shifts such as common corruptions, and adversarial
attacks. The existence of adversarial examples and data corruption
significantly reduces the performance of deep classification models.
Researchers have made strides in developing robust neural architectures to
bolster decisions of deep classifiers. However, most of these works rely on
effective adversarial training methods, and predominantly focus on overall
model robustness, disregarding class-wise differences in robustness, which are
critical. Exploiting weakly robust classes is a potential avenue for attackers
to fool the image recognition models. Therefore, this study investigates
class-to-class biases across adversarially trained robust classification models
to understand their latent space structures and analyze their strong and weak
class-wise properties. We further assess the robustness of classes against
common corruptions and adversarial attacks, recognizing that class
vulnerability extends beyond the number of correct classifications for a
specific class. We find that the number of false positives of classes as
specific target classes significantly impacts their vulnerability to attacks.
Through our analysis on the Class False Positive Score, we assess a fair
evaluation of how susceptible each class is to misclassification.

We introduce FaceGPT, a self-supervised learning framework for Large
Vision-Language Models (VLMs) to reason about 3D human faces from images and
text. Typical 3D face reconstruction methods are specialized algorithms that
lack semantic reasoning capabilities. FaceGPT overcomes this limitation by
embedding the parameters of a 3D morphable face model (3DMM) into the token
space of a VLM, enabling the generation of 3D faces from both textual and
visual inputs. FaceGPT is trained in a self-supervised manner as a model-based
autoencoder from in-the-wild images. In particular, the hidden state of LLM is
projected into 3DMM parameters and subsequently rendered as 2D face image to
guide the self-supervised learning process via image-based reconstruction.
Without relying on expensive 3D annotations of human faces, FaceGPT obtains a
detailed understanding about 3D human faces, while preserving the capacity to
understand general user instructions. Our experiments demonstrate that FaceGPT
not only achieves high-quality 3D face reconstructions but also retains the
ability for general-purpose visual instruction following. Furthermore, FaceGPT
learns fully self-supervised to generate 3D faces based on complex textual
inputs, which opens a new direction in human face analysis.

Synthesizing natural human motion that adapts to complex environments while
allowing creative control remains a fundamental challenge in motion synthesis.
Existing models often fall short, either by assuming flat terrain or lacking
the ability to control motion semantics through text. To address these
limitations, we introduce SCENIC, a diffusion model designed to generate human
motion that adapts to dynamic terrains within virtual scenes while enabling
semantic control through natural language. The key technical challenge lies in
simultaneously reasoning about complex scene geometry while maintaining text
control. This requires understanding both high-level navigation goals and
fine-grained environmental constraints. The model must ensure physical
plausibility and precise navigation across varied terrain, while also
preserving user-specified text control, such as ``carefully stepping over
obstacles" or ``walking upstairs like a zombie." Our solution introduces a
hierarchical scene reasoning approach. At its core is a novel scene-dependent,
goal-centric canonicalization that handles high-level goal constraint, and is
complemented by an ego-centric distance field that captures local geometric
details. This dual representation enables our model to generate physically
plausible motion across diverse 3D scenes. By implementing frame-wise text
alignment, our system achieves seamless transitions between different motion
styles while maintaining scene constraints. Experiments demonstrate our novel
diffusion model generates arbitrarily long human motions that both adapt to
complex scenes with varying terrain surfaces and respond to textual prompts.
Additionally, we show SCENIC can generalize to four real-scene datasets. Our
code, dataset, and models will be released at
\url{https://virtualhumans.mpi-inf.mpg.de/scenic/}.

Sampling-based decoding strategies have been widely adopted for Large
Language Models (LLMs) in numerous applications, which target a balance between
diversity and quality via temperature tuning and tail truncation (e.g., top-k
and top-p sampling). Considering the high dynamic range of the candidate
next-token given different prefixes, recent studies propose to adaptively
truncate the tail of LLM's predicted distribution. Although improved results
haven been reported with these methods on open-ended text generation tasks, the
results are highly dependent on the curated truncation parameters and exemplar
text. In this paper, we propose a systematic way to estimate the intrinsic
capacity of a truncation sampling method by considering the trade-off between
diversity and risk at each decoding step, based on our collected prefix tree
which preserves the context of a full sentence. Our work provides a
comprehensive comparison between existing truncation sampling methods, as well
as their recommended parameters as a guideline for users.

Deep learning models have proven to be successful in a wide
range of machine learning tasks. Yet, they are often highly sensitive to
perturbations on the input data which can lead to incorrect decisions
with high confidence, hampering their deployment for practical
use-cases. Thus, finding architectures that are (more) robust against
perturbations has received much attention in recent years. Just like the
search for well-performing architectures in terms of clean accuracy,
this usually involves a tedious trial-and-error process with one
additional challenge: the evaluation of a network's robustness is
significantly more expensive than its evaluation for clean accuracy.
Thus, the aim of this paper is to facilitate better streamlined research
on architectural design choices with respect to their impact on
robustness as well as, for example, the evaluation of surrogate measures
for robustness. We therefore borrow one of the most commonly considered
search spaces for neural architecture search for image classification,
NAS-Bench-201, which contains a manageable size of 6466 non-isomorphic
network designs. We evaluate all these networks on a range of common
adversarial attacks and corruption types and introduce a database on
neural architecture design and robustness evaluations. We further
present three exemplary use cases of this dataset, in which we (i)
benchmark robustness measurements based on Jacobian and Hessian matrices
for their robustness predictability, (ii) perform neural architecture
search on robust accuracies, and (iii) provide an initial analysis of
how architectural design choices affect robustness. We find that
carefully crafting the topology of a network can have substantial impact
on its robustness, where networks with the same parameter count range in
mean adversarial robust accuracy from 20%-41%.

Local coherence is essential for long-form text generation models. We identify two important aspects of local coherence within the visual storytelling task: (1) the model needs to represent re-occurrences of characters within the image sequence in order to mention them correctly in the story; (2) character representations should enable us to find instances of the same characters and distinguish different characters. In this paper, we propose a loss function inspired by a linguistic theory of coherence for self-supervised learning for image sequence representations. We further propose combining features from an object and a face detector to construct stronger character features. To evaluate input-output relevance that current reference-based metrics don't measure, we propose a character matching metric to check whether the models generate referring expressions correctly for characters in input image sequences. Experiments on a visual story generation dataset show that our proposed features and loss function are effective for generating more coherent and visually grounded stories.

Human performance capture is a highly important computer vision problem with
many applications in movie production and virtual/augmented reality. Many
previous performance capture approaches either required expensive multi-view
setups or did not recover dense space-time coherent geometry with
frame-to-frame correspondences. We propose a novel deep learning approach for
monocular dense human performance capture. Our method is trained in a weakly
supervised manner based on multi-view supervision completely removing the need
for training data with 3D ground truth annotations. The network architecture is
based on two separate networks that disentangle the task into a pose estimation
and a non-rigid surface deformation step. Extensive qualitative and
quantitative evaluations show that our approach outperforms the state of the
art in terms of quality and robustness. This work is an extended version of
DeepCap where we provide more detailed explanations, comparisons and results as
well as applications.

Minimum cost lifted multicut problem is a generalization of the multicut problem and is a means to optimizing a decomposition of a graph w.r.t. both positive and negative edge costs. Its main advantage is that multicut-based formulations do not require the number of components given a priori; instead, it is deduced from the solution. However, the standard multicut cost function is limited to pairwise relationships between nodes, while several important applications either require or can benefit from a higher-order cost function, i.e. hyper-edges. In this paper, we propose a pseudo-boolean formulation for a multiple model fitting problem. It is based on a formulation of any-order minimum cost lifted multicuts, which allows to partition an undirected graph with pairwise connectivity such as to minimize costs defined over any set of hyper-edges. As the proposed formulation is NP-hard and the branch-and-bound algorithm is too slow in practice, we propose an efficient local search algorithm for inference into resulting problems. We demonstrate versatility and effectiveness of our approach in several applications: geometric multiple model fitting, homography and motion estimation, motion segmentation.

Transformers increasingly dominate the machine learning landscape across many
tasks and domains, which increases the importance for understanding their
outputs. While their attention modules provide partial insight into their inner
workings, the attention scores have been shown to be insufficient for
explaining the models as a whole. To address this, we propose B-cos
transformers, which inherently provide holistic explanations for their
decisions. Specifically, we formulate each model component - such as the
multi-layer perceptrons, attention layers, and the tokenisation module - to be
dynamic linear, which allows us to faithfully summarise the entire transformer
via a single linear transform. We apply our proposed design to Vision
Transformers (ViTs) and show that the resulting models, dubbed Bcos-ViTs, are
highly interpretable and perform competitively to baseline ViTs on ImageNet.
Code will be made available soon.

Much of the world's most valued data is stored in relational databases and
data warehouses, where the data is organized into many tables connected by
primary-foreign key relations. However, building machine learning models using
this data is both challenging and time consuming. The core problem is that no
machine learning method is capable of learning on multiple tables
interconnected by primary-foreign key relations. Current methods can only learn
from a single table, so the data must first be manually joined and aggregated
into a single training table, the process known as feature engineering. Feature
engineering is slow, error prone and leads to suboptimal models. Here we
introduce an end-to-end deep representation learning approach to directly learn
on data laid out across multiple tables. We name our approach Relational Deep
Learning (RDL). The core idea is to view relational databases as a temporal,
heterogeneous graph, with a node for each row in each table, and edges
specified by primary-foreign key links. Message Passing Graph Neural Networks
can then automatically learn across the graph to extract representations that
leverage all input data, without any manual feature engineering. Relational
Deep Learning leads to more accurate models that can be built much faster. To
facilitate research in this area, we develop RelBench, a set of benchmark
datasets and an implementation of Relational Deep Learning. The data covers a
wide spectrum, from discussions on Stack Exchange to book reviews on the Amazon
Product Catalog. Overall, we define a new research area that generalizes graph
machine learning and broadens its applicability to a wide set of AI use cases.

Convolutional neural networks encode images through a sequence of
convolutions, normalizations and non-linearities as well as downsampling
operations into potentially strong semantic embeddings. Yet, previous work
showed that even slight mistakes during sampling, leading to aliasing, can be
directly attributed to the networks' lack in robustness. To address such issues
and facilitate simpler and faster adversarial training, [12] recently proposed
FLC pooling, a method for provably alias-free downsampling - in theory. In this
work, we conduct a further analysis through the lens of signal processing and
find that such current pooling methods, which address aliasing in the frequency
domain, are still prone to spectral leakage artifacts. Hence, we propose
aliasing and spectral artifact-free pooling, short ASAP. While only introducing
a few modifications to FLC pooling, networks using ASAP as downsampling method
exhibit higher native robustness against common corruptions, a property that
FLC pooling was missing. ASAP also increases native robustness against
adversarial attacks on high and low resolution data while maintaining similar
clean accuracy or even outperforming the baseline.

Zero-shot semantic segmentation (ZS3) aims to segment the novel categories
that have not been seen in the training. Existing works formulate ZS3 as a
pixel-level zero-shot classification problem, and transfer semantic knowledge
from seen classes to unseen ones with the help of language models pre-trained
only with texts. While simple, the pixel-level ZS3 formulation shows the
limited capability to integrate vision-language models that are often
pre-trained with image-text pairs and currently demonstrate great potential for
vision tasks. Inspired by the observation that humans often perform
segment-level semantic labeling, we propose to decouple the ZS3 into two
sub-tasks: 1) a class-agnostic grouping task to group the pixels into segments.
2) a zero-shot classification task on segments. The former sub-task does not
involve category information and can be directly transferred to group pixels
for unseen classes. The latter subtask performs at segment-level and provides a
natural way to leverage large-scale vision-language models pre-trained with
image-text pairs (e.g. CLIP) for ZS3. Based on the decoupling formulation, we
propose a simple and effective zero-shot semantic segmentation model, called
ZegFormer, which outperforms the previous methods on ZS3 standard benchmarks by
large margins, e.g., 35 points on the PASCAL VOC and 3 points on the COCO-Stuff
in terms of mIoU for unseen classes. Code will be released at
github.com/dingjiansw101/ZegFormer.

In this paper, we propose a novel co-learning framework (CoSSL) with
decoupled representation learning and classifier learning for imbalanced SSL.
To handle the data imbalance, we devise Tail-class Feature Enhancement (TFE)
for classifier learning. Furthermore, the current evaluation protocol for
imbalanced SSL focuses only on balanced test sets, which has limited
practicality in real-world scenarios. Therefore, we further conduct a
comprehensive evaluation under various shifted test distributions. In
experiments, we show that our approach outperforms other methods over a large
range of shifted distributions, achieving state-of-the-art performance on
benchmark datasets ranging from CIFAR-10, CIFAR-100, ImageNet, to Food-101. Our
code will be made publicly available.

As acquiring pixel-wise annotations of real-world images for semantic
segmentation is a costly process, a model can instead be trained with more
accessible synthetic data and adapted to real images without requiring their
annotations. This process is studied in unsupervised domain adaptation (UDA).
Even though a large number of methods propose new adaptation strategies, they
are mostly based on outdated network architectures. As the influence of recent
network architectures has not been systematically studied, we first benchmark
different network architectures for UDA and then propose a novel UDA method,
DAFormer, based on the benchmark results. The DAFormer network consists of a
Transformer encoder and a multi-level context-aware feature fusion decoder. It
is enabled by three simple but crucial training strategies to stabilize the
training and to avoid overfitting DAFormer to the source domain: While the Rare
Class Sampling on the source domain improves the quality of pseudo-labels by
mitigating the confirmation bias of self-training towards common classes, the
Thing-Class ImageNet Feature Distance and a learning rate warmup promote
feature transfer from ImageNet pretraining. DAFormer significantly improves the
state-of-the-art performance by 10.8 mIoU for GTA->Cityscapes and 5.4 mIoU for
Synthia->Cityscapes and enables learning even difficult classes such as train,
bus, and truck well. The implementation is available at
github.com/lhoyer/DAFormer.

Although considerable progress has been made in semantic scene understanding
under clear weather, it is still a tough problem under adverse weather
conditions, such as dense fog, due to the uncertainty caused by imperfect
observations. Besides, difficulties in collecting and labeling foggy images
hinder the progress of this field. Considering the success in semantic scene
understanding under clear weather, we think it is reasonable to transfer
knowledge learned from clear images to the foggy domain. As such, the problem
becomes to bridge the domain gap between clear images and foggy images. Unlike
previous methods that mainly focus on closing the domain gap caused by fog --
defogging the foggy images or fogging the clear images, we propose to alleviate
the domain gap by considering fog influence and style variation simultaneously.
The motivation is based on our finding that the style-related gap and the
fog-related gap can be divided and closed respectively, by adding an
intermediate domain. Thus, we propose a new pipeline to cumulatively adapt
style, fog and the dual-factor (style and fog). Specifically, we devise a
unified framework to disentangle the style factor and the fog factor
separately, and then the dual-factor from images in different domains.
Furthermore, we collaborate the disentanglement of three factors with a novel
cumulative loss to thoroughly disentangle these three factors. Our method
achieves the state-of-the-art performance on three benchmarks and shows
generalization ability in rainy and snowy scenes.

Multi-Camera Multi-Object Tracking is currently drawing attention in the
computer vision field due to its superior performance in real-world
applications such as video surveillance with crowded scenes or in vast space.
In this work, we propose a mathematically elegant multi-camera multiple object
tracking approach based on a spatial-temporal lifted multicut formulation. Our
model utilizes state-of-the-art tracklets produced by single-camera trackers as
proposals. As these tracklets may contain ID-Switch errors, we refine them
through a novel pre-clustering obtained from 3D geometry projections. As a
result, we derive a better tracking graph without ID switches and more precise
affinity costs for the data association phase. Tracklets are then matched to
multi-camera trajectories by solving a global lifted multicut formulation that
incorporates short and long-range temporal interactions on tracklets located in
the same camera as well as inter-camera ones. Experimental results on the
WildTrack dataset yield near-perfect result, outperforming state-of-the-art
trackers on Campus while being on par on the PETS-09 dataset. We will make our
implementations available upon acceptance of the paper.

3D human motion capture from monocular RGB images respecting interactions of
a subject with complex and possibly deformable environments is a very
challenging, ill-posed and under-explored problem. Existing methods address it
only weakly and do not model possible surface deformations often occurring when
humans interact with scene surfaces. In contrast, this paper proposes
MoCapDeform, i.e., a new framework for monocular 3D human motion capture that
is the first to explicitly model non-rigid deformations of a 3D scene for
improved 3D human pose estimation and deformable environment reconstruction.
MoCapDeform accepts a monocular RGB video and a 3D scene mesh aligned in the
camera space. It first localises a subject in the input monocular video along
with dense contact labels using a new raycasting based strategy. Next, our
human-environment interaction constraints are leveraged to jointly optimise
global 3D human poses and non-rigid surface deformations. MoCapDeform achieves
superior accuracy than competing methods on several datasets, including our
newly recorded one with deforming background scenes.

The minimum cost multicut problem is the NP-hard/APX-hard combinatorial
optimization problem of partitioning a real-valued edge-weighted graph such as
to minimize the total cost of the partition. While graph convolutional neural
networks (GNN) have proven to be promising in the context of combinatorial
optimization, most of them are only tailored to or tested on positive-valued
edge weights, i.e. they do not comply to the nature of the multicut problem. We
therefore adapt various GNN architectures including Graph Convolutional
Networks, Signed Graph Convolutional Networks and Graph Isomorphic Networks to
facilitate the efficient encoding of real-valued edge costs. Moreover, we
employ a reformulation of the multicut ILP constraints to a polynomial program
as loss function that allows to learn feasible multicut solutions in a scalable
way. Thus, we provide the first approach towards end-to-end trainable
multicuts. Our findings support that GNN approaches can produce good solutions
in practice while providing lower computation times and largely improved
scalability compared to LP solvers and optimized heuristics, especially when
considering large instances.

The Minimum Cost Multicut Problem (MP) is a popular way for obtaining a graph
decomposition by optimizing binary edge labels over edge costs. While the
formulation of a MP from independently estimated costs per edge is highly
flexible and intuitive, solving the MP is NP-hard and time-expensive. As a
remedy, recent work proposed to predict edge probabilities with awareness to
potential conflicts by incorporating cycle constraints in the prediction
process. We argue that such formulation, while providing a first step towards
end-to-end learnable edge weights, is suboptimal, since it is built upon a
loose relaxation of the MP. We therefore propose an adaptive CRF that allows to
progressively consider more violated constraints and, in consequence, to issue
solutions with higher validity. Experiments on the BSDS500 benchmark for
natural image segmentation as well as on electron microscopic recordings show
that our approach yields more precise edge detection and image segmentation.

Planning an optimal route in a complex environment requires efficient
reasoning about the surrounding scene. While human drivers prioritize important
objects and ignore details not relevant to the decision, learning-based
planners typically extract features from dense, high-dimensional grid
representations containing all vehicle and road context information. In this
paper, we propose PlanT, a novel approach for planning in the context of
self-driving that uses a standard transformer architecture. PlanT is based on
imitation learning with a compact object-level input representation. On the
Longest6 benchmark for CARLA, PlanT outperforms all prior methods (matching the
driving score of the expert) while being 5.3x faster than equivalent
pixel-based planning baselines during inference. Combining PlanT with an
off-the-shelf perception module provides a sensor-based driving system that is
more than 10 points better in terms of driving score than the existing state of
the art. Furthermore, we propose an evaluation protocol to quantify the ability
of planners to identify relevant objects, providing insights regarding their
decision-making. Our results indicate that PlanT can focus on the most relevant
object in the scene, even when this object is geometrically distant.

Semi-supervised learning (SSL) has shown great promise in leveraging
unlabeled data to improve model performance. While standard SSL assumes uniform
data distribution, we consider a more realistic and challenging setting called
imbalanced SSL, where imbalanced class distributions occur in both labeled and
unlabeled data. Although there are existing endeavors to tackle this challenge,
their performance degenerates when facing severe imbalance since they can not
reduce the class imbalance sufficiently and effectively. In this paper, we
study a simple yet overlooked baseline -- SimiS -- which tackles data imbalance
by simply supplementing labeled data with pseudo-labels, according to the
difference in class distribution from the most frequent class. Such a simple
baseline turns out to be highly effective in reducing class imbalance. It
outperforms existing methods by a significant margin, e.g., 12.8%, 13.6%, and
16.7% over previous SOTA on CIFAR100-LT, FOOD101-LT, and ImageNet127
respectively. The reduced imbalance results in faster convergence and better
pseudo-label accuracy of SimiS. The simplicity of our method also makes it
possible to be combined with other re-balancing techniques to improve the
performance further. Moreover, our method shows great robustness to a wide
range of data distributions, which holds enormous potential in practice. Code
will be publicly available.

Unintentional actions are rare occurrences that are difficult to define
precisely and that are highly dependent on the temporal context of the action.
In this work, we explore such actions and seek to identify the points in videos
where the actions transition from intentional to unintentional. We propose a
multi-stage framework that exploits inherent biases such as motion speed,
motion direction, and order to recognize unintentional actions. To enhance
representations via self-supervised training for the task of unintentional
action recognition we propose temporal transformations, called Temporal
Transformations of Inherent Biases of Unintentional Actions (T2IBUA). The
multi-stage approach models the temporal information on both the level of
individual frames and full clips. These enhanced representations show strong
performance for unintentional action recognition tasks. We provide an extensive
ablation study of our framework and report results that significantly improve
over the state-of-the-art.

Improving model's generalizability against domain shifts is crucial,
especially for safety-critical applications such as autonomous driving.
Real-world domain styles can vary substantially due to environment changes and
sensor noises, but deep models only know the training domain style. Such domain
style gap impedes model generalization on diverse real-world domains. Our
proposed Normalization Perturbation (NP) can effectively overcome this domain
style overfitting problem. We observe that this problem is mainly caused by the
biased distribution of low-level features learned in shallow CNN layers. Thus,
we propose to perturb the channel statistics of source domain features to
synthesize various latent styles, so that the trained deep model can perceive
diverse potential domains and generalizes well even without observations of
target domain data in training. We further explore the style-sensitive channels
for effective style synthesis. Normalization Perturbation only relies on a
single source domain and is surprisingly effective and extremely easy to
implement. Extensive experiments verify the effectiveness of our method for
generalizing models under real-world domain shifts.

In everyday lives, humans naturally modify the surrounding environment
through interactions, e.g., moving a chair to sit on it. To reproduce such
interactions in virtual spaces (e.g., metaverse), we need to be able to capture
and model them, including changes in the scene geometry, ideally from
ego-centric input alone (head camera and body-worn inertial sensors). This is
an extremely hard problem, especially since the object/scene might not be
visible from the head camera (e.g., a human not looking at a chair while
sitting down, or not looking at the door handle while opening a door). In this
paper, we present HOPS, the first method to capture interactions such as
dragging objects and opening doors from ego-centric data alone. Central to our
method is reasoning about human-object interactions, allowing to track objects
even when they are not visible from the head camera. HOPS localizes and
registers both the human and the dynamic object in a pre-scanned static scene.
HOPS is an important first step towards advanced AR/VR applications based on
immersive virtual universes, and can provide human-centric training data to
teach machines to interact with their surroundings. The supplementary video,
data, and code will be available on our project page at
virtualhumans.mpi-inf.mpg.de/hops/

This paper introduces DGNet, a novel deep framework that exploits object
gradient supervision for camouflaged object detection (COD). It decouples the
task into two connected branches, i.e., a context and a texture encoder. The
essential connection is the gradient-induced transition, representing a soft
grouping between context and texture features. Benefiting from the simple but
efficient framework, DGNet outperforms existing state-of-the-art COD models by
a large margin. Notably, our efficient version, DGNet-S, runs in real-time (80
fps) and achieves comparable results to the cutting-edge model
JCSOD-CVPR$_{21}$ with only 6.82% parameters. Application results also show
that the proposed DGNet performs well in polyp segmentation, defect detection,
and transparent object segmentation tasks. Codes will be made available at
github.com/GewelsJI/DGNet.

In this report, we present the 1st place solution for motion prediction track
in 2022 Waymo Open Dataset Challenges. We propose a novel Motion Transformer
framework for multimodal motion prediction, which introduces a small set of
novel motion query pairs for generating better multimodal future trajectories
by jointly performing the intention localization and iterative motion
refinement. A simple model ensemble strategy with non-maximum-suppression is
adopted to further boost the final performance. Our approach achieves the 1st
place on the motion prediction leaderboard of 2022 Waymo Open Dataset
Challenges, outperforming other methods with remarkable margins. Code will be
available at github.com/sshaoshuai/MTR.

Deep learning is becoming increasingly relevant for many high-stakes applications such as autonomous driving or medical diagnosis where wrong decisions can have massive impact on human lives. Unfortunately, deep neural networks are typically assessed solely based on generalization, e.g., accuracy on a fixed test set. However, this is clearly insufficient for safe deployment as potential malicious actors and distribution shifts or the effects of quantization and unreliable hardware are disregarded. Thus, recent work additionally evaluates performance on potentially manipulated or corrupted inputs as well as after quantization and deployment on specialized hardware. In such settings, it is also important to obtain reasonable estimates of the model's confidence alongside its predictions. This thesis studies robustness and uncertainty estimation in deep learning along three main directions: First, we consider so-called adversarial examples, slightly perturbed inputs causing severe drops in accuracy. Second, we study weight perturbations, focusing particularly on bit errors in quantized weights. This is relevant for deploying models on special-purpose hardware for efficient inference, so-called accelerators. Finally, we address uncertainty estimation to improve robustness and provide meaningful statistical performance guarantees for safe deployment. In detail, we study the existence of adversarial examples with respect to the underlying data manifold. In this context, we also investigate adversarial training which improves robustness by augmenting training with adversarial examples at the cost of reduced accuracy. We show that regular adversarial examples leave the data manifold in an almost orthogonal direction. While we find no inherent trade-off between robustness and accuracy, this contributes to a higher sample complexity as well as severe overfitting of adversarial training. Using a novel measure of flatness in the robust loss landscape with respect to weight changes, we also show that robust overfitting is caused by converging to particularly sharp minima. In fact, we find a clear correlation between flatness and good robust generalization. Further, we study random and adversarial bit errors in quantized weights. In accelerators, random bit errors occur in the memory when reducing voltage with the goal of improving energy-efficiency. Here, we consider a robust quantization scheme, use weight clipping as regularization and perform random bit error training to improve bit error robustness, allowing considerable energy savings without requiring hardware changes. In contrast, adversarial bit errors are maliciously introduced through hardware- or software-based attacks on the memory, with severe consequences on performance. We propose a novel adversarial bit error attack to study this threat and use adversarial bit error training to improve robustness and thereby also the accelerator's security. Finally, we view robustness in the context of uncertainty estimation. By encouraging low-confidence predictions on adversarial examples, our confidence-calibrated adversarial training successfully rejects adversarial, corrupted as well as out-of-distribution examples at test time. Thereby, we are also able to improve the robustness-accuracy trade-off compared to regular adversarial training. However, even robust models do not provide any guarantee for safe deployment. To address this problem, conformal prediction allows the model to predict confidence sets with user-specified guarantee of including the true label. Unfortunately, as conformal prediction is usually applied after training, the model is trained without taking this calibration step into account. To address this limitation, we propose conformal training which allows training conformal predictors end-to-end with the underlying model. This not only improves the obtained uncertainty estimates but also enables optimizing application-specific objectives without losing the provided guarantee. Besides our work on robustness or uncertainty, we also address the problem of 3D shape completion of partially observed point clouds. Specifically, we consider an autonomous driving or robotics setting where vehicles are commonly equipped with LiDAR or depth sensors and obtaining a complete 3D representation of the environment is crucial. However, ground truth shapes that are essential for applying deep learning techniques are extremely difficult to obtain. Thus, we propose a weakly-supervised approach that can be trained on the incomplete point clouds while offering efficient inference. In summary, this thesis contributes to our understanding of robustness against both input and weight perturbations. To this end, we also develop methods to improve robustness alongside uncertainty estimation for safe deployment of deep learning methods in high-stakes applications. In the particular context of autonomous driving, we also address 3D shape completion of sparse point clouds.

Structured prediction problems are one of the fundamental tools in machine
learning. In order to facilitate algorithm development for their numerical
solution, we collect in one place a large number of datasets in easy to read
formats for a diverse set of problem classes. We provide archival links to
datasets, description of the considered problems and problem formats, and a
short summary of problem characteristics including size, number of instances
etc. For reference we also give a non-exhaustive selection of algorithms
proposed in the literature for their solution. We hope that this central
repository will make benchmarking and comparison to established works easier.
We welcome submission of interesting new datasets and algorithms for inclusion
in our archive.

Modern deep learning architecture utilize batch normalization (BN) to
stabilize training and improve accuracy. It has been shown that the BN layers
alone are surprisingly expressive. In the context of robustness against
adversarial examples, however, BN is argued to increase vulnerability. That is,
BN helps to learn fragile features. Nevertheless, BN is still used in
adversarial training, which is the de-facto standard to learn robust features.
In order to shed light on the role of BN in adversarial training, we
investigate to what extent the expressiveness of BN can be used to robustify
fragile features in comparison to random features. On CIFAR10, we find that
adversarially fine-tuning just the BN layers can result in non-trivial
adversarial robustness. Adversarially training only the BN layers from scratch,
in contrast, is not able to convey meaningful adversarial robustness. Our
results indicate that fragile features can be used to learn models with
moderate adversarial robustness, while random features cannot

Current efficient LiDAR-based detection frameworks are lacking in exploiting
object relations, which naturally present in both spatial and temporal manners.
To this end, we introduce a simple, efficient, and effective two-stage
detector, termed as Ret3D. At the core of Ret3D is the utilization of novel
intra-frame and inter-frame relation modules to capture the spatial and
temporal relations accordingly. More Specifically, intra-frame relation module
(IntraRM) encapsulates the intra-frame objects into a sparse graph and thus
allows us to refine the object features through efficient message passing. On
the other hand, inter-frame relation module (InterRM) densely connects each
object in its corresponding tracked sequences dynamically, and leverages such
temporal information to further enhance its representations efficiently through
a lightweight transformer network. We instantiate our novel designs of IntraRM
and InterRM with general center-based or anchor-based detectors and evaluate
them on Waymo Open Dataset (WOD). With negligible extra overhead, Ret3D
achieves the state-of-the-art performance, being 5.5% and 3.2% higher than the
recent competitor in terms of the LEVEL 1 and LEVEL 2 mAPH metrics on vehicle
detection, respectively.

We present TOCH, a method for refining incorrect 3D hand-object interaction
sequences using a data prior. Existing hand trackers, especially those that
rely on very few cameras, often produce visually unrealistic results with
hand-object intersection or missing contacts. Although correcting such errors
requires reasoning about temporal aspects of interaction, most previous work
focus on static grasps and contacts. The core of our method are TOCH fields, a
novel spatio-temporal representation for modeling correspondences between hands
and objects during interaction. The key component is a point-wise
object-centric representation which encodes the hand position relative to the
object. Leveraging this novel representation, we learn a latent manifold of
plausible TOCH fields with a temporal denoising auto-encoder. Experiments
demonstrate that TOCH outperforms state-of-the-art (SOTA) 3D hand-object
interaction models, which are limited to static grasps and contacts. More
importantly, our method produces smooth interactions even before and after
contact. Using a single trained TOCH model, we quantitatively and qualitatively
demonstrate its usefulness for 1) correcting erroneous reconstruction results
from off-the-shelf RGB/RGB-D hand-object reconstruction methods, 2) de-noising,
and 3) grasp transfer across objects. We will release our code and trained
model on our project page at virtualhumans.mpi-inf.mpg.de/toch/

Skeleton-based action recognition aims to predict human actions given human
joint coordinates with skeletal interconnections. To model such off-grid data
points and their co-occurrences, Transformer-based formulations would be a
natural choice. However, Transformers still lag behind state-of-the-art methods
using graph convolutional networks (GCNs). Transformers assume that the input
is permutation-invariant and homogeneous (partially alleviated by positional
encoding), which ignores an important characteristic of skeleton data, i.e.,
bone connectivity. Furthermore, each type of body joint has a clear physical
meaning in human motion, i.e., motion retains an intrinsic relationship
regardless of the joint coordinates, which is not explored in Transformers. In
fact, certain re-occurring groups of body joints are often involved in specific
actions, such as the subconscious hand movement for keeping balance. Vanilla
attention is incapable of describing such underlying relations that are
persistent and beyond pair-wise. In this work, we aim to exploit these unique
aspects of skeleton data to close the performance gap between Transformers and
GCNs. Specifically, we propose a new self-attention (SA) extension, named
Hypergraph Self-Attention (HyperSA), to incorporate inherently higher-order
relations into the model. The K-hop relative positional embeddings are also
employed to take bone connectivity into account. We name the resulting model
Hyperformer, and it achieves comparable or better performance w.r.t. accuracy
and efficiency than state-of-the-art GCN architectures on NTU RGB+D, NTU RGB+D
120, and Northwestern-UCLA datasets. On the largest NTU RGB+D 120 dataset, the
significantly improved performance reached by our Hyperformer demonstrates the
underestimated potential of Transformer models in this field.

Deep neural network (DNN) accelerators received considerable attention in
past years due to saved energy compared to mainstream hardware. Low-voltage
operation of DNN accelerators allows to further reduce energy consumption
significantly, however, causes bit-level failures in the memory storing the
quantized DNN weights. In this paper, we show that a combination of robust
fixed-point quantization, weight clipping, and random bit error training
(RandBET) improves robustness against random bit errors in (quantized) DNN
weights significantly. This leads to high energy savings from both low-voltage
operation as well as low-precision quantization. Our approach generalizes
across operating voltages and accelerators, as demonstrated on bit errors from
profiled SRAM arrays. We also discuss why weight clipping alone is already a
quite effective way to achieve robustness against bit errors. Moreover, we
specifically discuss the involved trade-offs regarding accuracy, robustness and
precision: Without losing more than 1% in accuracy compared to a normally
trained 8-bit DNN, we can reduce energy consumption on CIFAR-10 by 20%. Higher
energy savings of, e.g., 30%, are possible at the cost of 2.5% accuracy, even
for 4-bit DNNs.

We present a massively parallel Lagrange decomposition method for solving 0-1
integer linear programs occurring in structured prediction. We propose a new
iterative update scheme for solving the Lagrangean dual and a perturbation
technique for decoding primal solutions. For representing subproblems we follow
Lange et al. (2021) and use binary decision diagrams (BDDs). Our primal and
dual algorithms require little synchronization between subproblems and
optimization over BDDs needs only elementary operations without complicated
control flow. This allows us to exploit the parallelism offered by GPUs for all
components of our method. We present experimental results on combinatorial
problems from MAP inference for Markov Random Fields, quadratic assignment and
cell tracking for developmental biology. Our highly parallel GPU implementation
improves upon the running times of the algorithms from Lange et al. (2021) by
up to an order of magnitude. In particular, we come close to or outperform some
state-of-the-art specialized heuristics while being problem agnostic.

Explaining the decision of a multi-modal decision-maker requires to determine
the evidence from both modalities. Recent advances in XAI provide explanations
for models trained on still images. However, when it comes to modeling multiple
sensory modalities in a dynamic world, it remains underexplored how to
demystify the mysterious dynamics of a complex multi-modal model. In this work,
we take a crucial step forward and explore learnable explanations for
audio-visual recognition. Specifically, we propose a novel space-time attention
network that uncovers the synergistic dynamics of audio and visual data over
both space and time. Our model is capable of predicting the audio-visual video
events, while justifying its decision by localizing where the relevant visual
cues appear, and when the predicted sounds occur in videos. We benchmark our
model on three audio-visual video event datasets, comparing extensively to
multiple recent multi-modal representation learners and intrinsic explanation
models. Experimental results demonstrate the clear superior performance of our
model over the existing methods on audio-visual video event recognition.
Moreover, we conduct an in-depth study to analyze the explainability of our
model based on robustness analysis via perturbation tests and pointing games
using human annotations.

Traditional domain adaptation addresses the task of adapting a model to a
novel target domain under limited or no additional supervision. While tackling
the input domain gap, the standard domain adaptation settings assume no domain
change in the output space. In semantic prediction tasks, different datasets
are often labeled according to different semantic taxonomies. In many
real-world settings, the target domain task requires a different taxonomy than
the one imposed by the source domain. We therefore introduce the more general
taxonomy adaptive domain adaptation (TADA) problem, allowing for inconsistent
taxonomies between the two domains. We further propose an approach that jointly
addresses the image-level and label-level domain adaptation. On the
label-level, we employ a bilateral mixed sampling strategy to augment the
target domain, and a relabelling method to unify and align the label spaces. We
address the image-level domain gap by proposing an uncertainty-rectified
contrastive learning method, leading to more domain-invariant and class
discriminative features. We extensively evaluate the effectiveness of our
framework under different TADA settings: open taxonomy, coarse-to-fine
taxonomy, and partially-overlapping taxonomy. Our framework outperforms
previous state-of-the-art by a large margin, while capable of adapting to
target taxonomies.

Compositional Zero-Shot learning (CZSL) aims to recognize unseen compositions
of state and object visual primitives seen during training. A problem with
standard CZSL is the assumption of knowing which unseen compositions will be
available at test time. In this work, we overcome this assumption operating on
the open world setting, where no limit is imposed on the compositional space at
test time, and the search space contains a large number of unseen compositions.
To address this problem, we propose a new approach, Compositional Cosine Graph
Embeddings (Co-CGE), based on two principles. First, Co-CGE models the
dependency between states, objects and their compositions through a graph
convolutional neural network. The graph propagates information from seen to
unseen concepts, improving their representations. Second, since not all unseen
compositions are equally feasible, and less feasible ones may damage the
learned representations, Co-CGE estimates a feasibility score for each unseen
composition, using the scores as margins in a cosine similarity-based loss and
as weights in the adjacency matrix of the graphs. Experiments show that our
approach achieves state-of-the-art performances in standard CZSL while
outperforming previous methods in the open world scenario.

Abstract
Humans are at the centre of a significant amount of research in computer vision.
Endowing machines with the ability to perceive people from visual data is an immense
scientific challenge with a high degree of direct practical relevance. Success in automatic
perception can be measured at different levels of abstraction, and this will depend on
which intelligent behaviour we are trying to replicate: the ability to localise persons in
an image or in the environment, understanding how persons are moving at the skeleton
and at the surface level, interpreting their interactions with the environment including
with other people, and perhaps even anticipating future actions. In this thesis we tackle
different sub-problems of the broad research area referred to as "looking at people",
aiming to perceive humans in images at different levels of granularity.
We start with bounding box-level pedestrian detection: We present a retrospective
analysis of methods published in the decade preceding our work, identifying various
strands of research that have advanced the state of the art. With quantitative exper-
iments, we demonstrate the critical role of developing better feature representations
and having the right training distribution. We then contribute two methods based
on the insights derived from our analysis: one that combines the strongest aspects of
past detectors and another that focuses purely on learning representations. The latter
method outperforms more complicated approaches, especially those based on hand-
crafted features. We conclude our work on pedestrian detection with a forward-looking
analysis that maps out potential avenues for future research.
We then turn to pixel-level methods: Perceiving humans requires us to both separate
them precisely from the background and identify their surroundings. To this end, we
introduce Cityscapes, a large-scale dataset for street scene understanding. This has since
established itself as a go-to benchmark for segmentation and detection. We additionally
develop methods that relax the requirement for expensive pixel-level annotations, focusing
on the task of boundary detection, i.e. identifying the outlines of relevant objects and
surfaces. Next, we make the jump from pixels to 3D surfaces, from localising and
labelling to fine-grained spatial understanding. We contribute a method for recovering
3D human shape and pose, which marries the advantages of learning-based and model-
based approaches.
We conclude the thesis with a detailed discussion of benchmarking practices in
computer vision. Among other things, we argue that the design of future datasets
should be driven by the general goal of combinatorial robustness besides task-specific
considerations.

Most learning methods for 3D data (point clouds, meshes) suffer significant
performance drops when the data is not carefully aligned to a canonical
orientation. Aligning real world 3D data collected from different sources is
non-trivial and requires manual intervention. In this paper, we propose the
Adjoint Rigid Transform (ART) Network, a neural module which can be integrated
with a variety of 3D networks to significantly boost their performance. ART
learns to rotate input shapes to a learned canonical orientation, which is
crucial for a lot of tasks such as shape reconstruction, interpolation,
non-rigid registration, and latent disentanglement. ART achieves this with
self-supervision and a rotation equivariance constraint on predicted rotations.
The remarkable result is that with only self-supervision, ART facilitates
learning a unique canonical orientation for both rigid and nonrigid shapes,
which leads to a notable boost in performance of aforementioned tasks. We will
release our code and pre-trained models for further research.

Modern approaches to pose and body shape estimation have recently achieved
strong performance even under challenging real-world conditions. Even from a
single image of a clothed person, a realistic looking body shape can be
inferred that captures a users' weight group and body shape type well. This
opens up a whole spectrum of applications -- in particular in fashion -- where
virtual try-on and recommendation systems can make use of these new and
automatized cues. However, a realistic depiction of the undressed body is
regarded highly private and therefore might not be consented by most people.
Hence, we ask if the automatic extraction of such information can be
effectively evaded. While adversarial perturbations have been shown to be
effective for manipulating the output of machine learning models -- in
particular, end-to-end deep learning approaches -- state of the art shape
estimation methods are composed of multiple stages. We perform the first
investigation of different strategies that can be used to effectively
manipulate the automatic shape estimation while preserving the overall
appearance of the original image.

Current datasets for video-based person re-identification (re-ID) do not
include structural knowledge in form of human pose annotations for the persons
of interest. Nonetheless, pose information is very helpful to disentangle
useful feature information from background or occlusion noise. Especially
real-world scenarios, such as surveillance, contain a lot of occlusions in
human crowds or by obstacles. On the other hand, video-based person re-ID can
benefit other tasks such as multi-person pose tracking in terms of robust
feature matching. For that reason, we present PoseTrackReID, a large-scale
dataset for multi-person pose tracking and video-based person re-ID. With
PoseTrackReID, we want to bridge the gap between person re-ID and multi-person
pose tracking. Additionally, this dataset provides a good benchmark for current
state-of-the-art methods on multi-frame person re-ID.

Intuitively, image classification should profit from using spatial
information. Recent work, however, suggests that this might be overrated in
standard CNNs. In this paper, we are pushing the envelope and aim to further
investigate the reliance on spatial information. We propose spatial shuffling
and GAP+FC to destroy spatial information during both training and testing
phases. Interestingly, we observe that spatial information can be deleted from
later layers with small performance drops, which indicates spatial information
at later layers is not necessary for good performance. For example, test
accuracy of VGG-16 only drops by 0.03% and 2.66% with spatial information
completely removed from the last 30% and 53% layers on CIFAR100, respectively.
Evaluation on several object recognition datasets (CIFAR100, Small-ImageNet,
ImageNet) with a wide range of CNN architectures (VGG16, ResNet50, ResNet152)
shows an overall consistent pattern.

Modern deep learning has enabled amazing developments of computer vision in recent years (Hinton and Salakhutdinov, 2006; Krizhevsky et al., 2012). As a fundamental task, semantic segmentation aims to predict class labels for each pixel of images, which empowers machines perception of the visual world. In spite of recent successes of fully convolutional networks (Long etal., 2015), several challenges remain to be addressed. In this thesis, we focus on this topic, under different kinds of input formats and various types of scenes. Specifically, our study contains two aspects: (1) Data-driven neural modules for improved performance. (2) Leverage of datasets w.r.t.training systems with higher performances and better data privacy guarantees. In the first part of this thesis, we improve semantic segmentation by designing new modules which are compatible with existing architectures. First, we develop a spatio-temporal data-driven pooling, which brings additional information of data (i.e. superpixels) into neural networks, benefiting the training of neural networks as well as the inference on novel data. We investigate our approach in RGB-D videos for segmenting indoor scenes, where depth provides complementary cues to colors and our model performs particularly well. Second, we design learnable dilated convolutions, which are the extension of standard dilated convolutions, whose dilation factors (Yu and Koltun, 2016) need to be carefully determined by hand to obtain decent performance. We present a method to learn dilation factors together with filter weights of convolutions to avoid a complicated search of dilation factors. We explore extensive studies on challenging street scenes, across various baselines with different complexity as well as several datasets at varying image resolutions. In the second part, we investigate how to utilize expensive training data. First, we start from the generative modelling and study the network architectures and the learning pipeline for generating multiple examples. We aim to improve the diversity of generated examples but also to preserve the comparable quality of the examples. Second, we develop a generative model for synthesizing features of a network. With a mixture of real images and synthetic features, we are able to train a segmentation model with better generalization capability. Our approach is evaluated on different scene parsing tasks to demonstrate the effectiveness of the proposed method. Finally, we study membership inference on the semantic segmentation task. We propose the first membership inference attack system against black-box semantic segmentation models, that tries to infer if a data pair is used as training data or not. From our observations, information on training data is indeed leaking. To mitigate the leakage, we leverage our synthetic features to perform prediction obfuscations, reducing the posterior distribution gaps between a training and a testing set. Consequently, our study provides not only an approach for detecting illegal use of data, but also the foundations for a safer use of semantic segmentation models.

Automatic detection of emergent leaders in small groups from nonverbal
behaviour is a growing research topic in social signal processing but existing
methods were evaluated on single datasets -- an unrealistic assumption for
real-world applications in which systems are required to also work in settings
unseen at training time. It therefore remains unclear whether current methods
for emergent leadership detection generalise to similar but new settings and to
which extent. To overcome this limitation, we are the first to study a
cross-dataset evaluation setting for the emergent leadership detection task. We
provide evaluations for within- and cross-dataset prediction using two current
datasets (PAVIS and MPIIGroupInteraction), as well as an investigation on the
robustness of commonly used feature channels (visual focus of attention, body
pose, facial action units, speaking activity) and online prediction in the
cross-dataset setting. Our evaluations show that using pose and eye contact
based features, cross-dataset prediction is possible with an accuracy of 0.68,
as such providing another important piece of the puzzle towards emergent
leadership detection in the real world.

When labeled training data is scarce, a promising data augmentation approach is to generate visual features of unknown classes using their attributes. To learn the class conditional distribution of CNN features, these models rely on pairs of image features and class attributes. Hence, they can not make use of the abundance of unlabeled data samples. In this paper, we tackle any-shot learning problems i.e. zero-shot and few-shot, in a unified feature generating framework that operates in both inductive and transductive learning settings. We develop a conditional generative model that combines the strength of VAE and GANs and in addition, via an unconditional discriminator, learns the marginal feature distribution of unlabeled images. We empirically show that our model learns highly discriminative CNN features for five datasets, i.e. CUB, SUN, AWA and ImageNet, and establish a new state-of-the-art in any-shot learning, i.e. inductive and transductive (generalized) zero- and few-shot learning settings. We also demonstrate that our learned features are interpretable: we visualize them by inverting them back to the pixel space and we explain them by generating textual arguments of why they are associated with a certain label.

When labeled training data is scarce, a promising data augmentation approach is to generate visual features of unknown classes using their attributes. To learn the class conditional distribution of CNN features, these models rely on pairs of image features and class attributes. Hence, they can not make use of the abundance of unlabeled data samples. In this paper, we tackle any-shot learning problems i.e. zero-shot and few-shot, in a unified feature generating framework that operates in both inductive and transductive learning settings. We develop a conditional generative model that combines the strength of VAE and GANs and in addition, via an unconditional discriminator, learns the marginal feature distribution of unlabeled images. We empirically show that our model learns highly discriminative CNN features for five datasets, i.e. CUB, SUN, AWA and ImageNet, and establish a new state-of-the-art in any-shot learning, i.e. inductive and transductive (generalized) zero- and few-shot learning settings. We also demonstrate that our learned features are interpretable: we visualize them by inverting them back to the pixel space and we explain them by generating textual arguments of why they are associated with a certain label.

Due to the importance of zero-shot learning, i.e. classifying images where
there is a lack of labeled training data, the number of proposed approaches has
recently increased steadily. We argue that it is time to take a step back and
to analyze the status quo of the area. The purpose of this paper is three-fold.
First, given the fact that there is no agreed upon zero-shot learning
benchmark, we first define a new benchmark by unifying both the evaluation
protocols and data splits of publicly available datasets used for this task.
This is an important contribution as published results are often not comparable
and sometimes even flawed due to, e.g. pre-training on zero-shot test classes.
Moreover, we propose a new zero-shot learning dataset, the Animals with
Attributes 2 (AWA2) dataset which we make publicly available both in terms of
image features and the images themselves. Second, we compare and analyze a
significant number of the state-of-the-art methods in depth, both in the
classic zero-shot setting but also in the more realistic generalized zero-shot
setting. Finally, we discuss in detail the limitations of the current status of
the area which can be taken as a basis for advancing it.

We present a simple yet effective method to infer detailed full human body
shape from only a single photograph. Our model can infer full-body shape
including face, hair, and clothing including wrinkles at interactive
frame-rates. Results feature details even on parts that are occluded in the
input image. Our main idea is to turn shape regression into an aligned
image-to-image translation problem. The input to our method is a partial
texture map of the visible region obtained from off-the-shelf methods. From a
partial texture, we estimate detailed normal and vector displacement maps,
which can be applied to a low-resolution smooth body model to add detail and
clothing. Despite being trained purely with synthetic data, our model
generalizes well to real-world photographs. Numerous results demonstrate the
versatility and robustness of our method.

Internal thought refers to the process of directing attention away from a
primary visual task to internal cognitive processing. Internal thought is a
pervasive mental activity and closely related to primary task performance. As
such, automatic detection of internal thought has significant potential for
user modelling in intelligent interfaces, particularly for e-learning
applications. Despite the close link between the eyes and the human mind, only
a few studies have investigated vergence behaviour during internal thought and
none has studied moment-to-moment detection of internal thought from gaze.
While prior studies relied on long-term data analysis and required a large
number of gaze characteristics, we describe a novel method that is
computationally light-weight and that only requires eye vergence information
that is readily available from binocular eye trackers. We further propose a
novel paradigm to obtain ground truth internal thought annotations that
exploits human blur perception. We evaluate our method for three increasingly
challenging detection tasks: (1) during a controlled math-solving task, (2)
during natural viewing of lecture videos, and (3) during daily activities, such
as coding, browsing, and reading. Results from these evaluations demonstrate
the performance and robustness of vergence-based detection of internal thought
and, as such, open up new directions for research on interfaces that adapt to
shifts of mental attention.

Recent methods to automatically calibrate stationary eye trackers were shown
to effectively reduce inherent calibration distortion. However, these methods
require additional information, such as mouse clicks or on-screen content. We
propose the first method that only requires users' eye movements to reduce
calibration distortion in the background while users naturally look at an
interface. Our method exploits that calibration distortion makes straight
saccade trajectories appear curved between the saccadic start and end points.
We show that this curving effect is systematic and the result of distorted gaze
projection plane. To mitigate calibration distortion, our method undistorts
this plane by straightening saccade trajectories using image warping. We show
that this approach improves over the common six-point calibration and is
promising for reducing distortion. As such, it provides a non-intrusive
solution to alleviating accuracy decrease of eye tracker during long-term use.

We consider general discrete Markov Random Fields(MRFs) with additional
bottleneck potentials which penalize the maximum (instead of the sum) over
local potential value taken by the MRF-assignment. Bottleneck potentials or
analogous constructions have been considered in (i) combinatorial optimization
(e.g. bottleneck shortest path problem, the minimum bottleneck spanning tree
problem, bottleneck function minimization in greedoids), (ii) inverse problems
with $L_{\infty}$-norm regularization, and (iii) valued constraint satisfaction
on the $(\min,\max)$-pre-semirings. Bottleneck potentials for general discrete
MRFs are a natural generalization of the above direction of modeling work to
Maximum-A-Posteriori (MAP) inference in MRFs. To this end, we propose MRFs
whose objective consists of two parts: terms that factorize according to (i)
$(\min,+)$, i.e. potentials as in plain MRFs, and (ii) $(\min,\max)$, i.e.
bottleneck potentials. To solve the ensuing inference problem, we propose
high-quality relaxations and efficient algorithms for solving them. We
empirically show efficacy of our approach on large scale seismic horizon
tracking problems.

Generative Adversarial Networks (GANs) can achieve state-of-the-art sample
quality in generative modelling tasks but suffer from the mode collapse
problem. Variational Autoencoders (VAE) on the other hand explicitly maximize a
reconstruction-based data log-likelihood forcing it to cover all modes, but
suffer from poorer sample quality. Recent works have proposed hybrid VAE-GAN
frameworks which integrate a GAN-based synthetic likelihood to the VAE
objective to address both the mode collapse and sample quality issues, with
limited success. This is because the VAE objective forces a trade-off between
the data log-likelihood and divergence to the latent prior. The synthetic
likelihood ratio term also shows instability during training. We propose a
novel objective with a "Best-of-Many-Samples" reconstruction cost and a stable
direct estimate of the synthetic likelihood. This enables our hybrid VAE-GAN
framework to achieve high data log-likelihood and low divergence to the latent
prior at the same time and shows significant improvement over both hybrid
VAE-GANS and plain GANs in mode coverage and quality.

Meta-learning has been shown to be an effective strategy for few-shot
learning. The key idea is to leverage a large number of similar few-shot tasks
in order to meta-learn how to best initiate a (single) base-learner for novel
few-shot tasks. While meta-learning how to initialize a base-learner has shown
promising results, it is well known that hyperparameter settings such as the
learning rate and the weighting of the regularization term are important to
achieve best performance. We thus propose to also meta-learn these
hyperparameters and in fact learn a time- and layer-varying scheme for learning
a base-learner on novel tasks. Additionally, we propose to learn not only a
single base-learner but an ensemble of several base-learners to obtain more
robust results. While ensembles of learners have shown to improve performance
in various settings, this is challenging for few-shot learning tasks due to the
limited number of training samples. Therefore, our approach also aims to
meta-learn how to effectively combine several base-learners. We conduct
extensive experiments and report top performance for five-class few-shot
recognition tasks on two challenging benchmarks: miniImageNet and
Fewshot-CIFAR100 (FC100).

Understanding physical phenomena is a key competence that enables humans and
animals to act and interact under uncertain perception in previously unseen
environments containing novel objects and their configurations. In this work,
we consider the problem of autonomous block stacking and explore solutions to
learning manipulation under physics constraints with visual perception inherent
to the task. Inspired by the intuitive physics in humans, we first present an
end-to-end learning-based approach to predict stability directly from
appearance, contrasting a more traditional model-based approach with explicit
3D representations and physical simulation. We study the model's behavior
together with an accompanied human subject test. It is then integrated into a
real-world robotic system to guide the placement of a single wood block into
the scene without collapsing existing tower structure. To further automate the
process of consecutive blocks stacking, we present an alternative approach
where the model learns the physics constraint through the interaction with the
environment, bypassing the dedicated physics learning as in the former part of
this work. In particular, we are interested in the type of tasks that require
the agent to reach a given goal state that may be different for every new
trial. Thereby we propose a deep reinforcement learning framework that learns
policies for stacking tasks which are parametrized by a target structure.

Today's deep learning systems deliver high performance based on end-to-end
training. While they deliver strong performance, these systems are hard to
interpret. To address this issue, we propose Semantic Bottleneck Networks
(SBN): deep networks with semantically interpretable intermediate layers that
all downstream results are based on. As a consequence, the analysis on what the
final prediction is based on is transparent to the engineer and failure cases
and modes can be analyzed and avoided by high-level reasoning. We present a
case study on street scene segmentation to demonstrate the feasibility and
power of SBN. In particular, we start from a well performing classic deep
network which we adapt to house a SB-Layer containing task related semantic
concepts (such as object-parts and materials). Importantly, we can recover
state of the art performance despite a drastic dimensionality reduction from
1000s (non-semantic feature) to 10s (semantic concept) channels. Additionally
we show how the activations of the SB-Layer can be used for both the
interpretation of failure cases of the network as well as for confidence
prediction of the resulting output. For the first time, e.g., we show
interpretable segmentation results for most predictions at over 99% accuracy.

Image-to-image (I2I) translation is a pixel-level mapping that requires a
large number of paired training data and often suffers from the problems of
high diversity and strong category bias in image scenes. In order to tackle
these problems, we propose a novel BiLevel (BiL) learning paradigm that
alternates the learning of two models, respectively at an instance-specific
(IS) and a general-purpose (GP) level. In each scene, the IS model learns to
maintain the specific scene attributes. It is initialized by the GP model that
learns from all the scenes to obtain the generalizable translation knowledge.
This GP initialization gives the IS model an efficient starting point, thus
enabling its fast adaptation to the new scene with scarce training data. We
conduct extensive I2I translation experiments on human face and street view
datasets. Quantitative results validate that our approach can significantly
boost the performance of classical I2I translation models, such as PG2 and
Pix2Pix. Our visualization results show both higher image quality and more
appropriate instance-specific details, e.g., the translated image of a person
looks more like that person in terms of identity.

We present a real-time approach for multi-person 3D motion capture at over 30
fps using a single RGB camera. It operates in generic scenes and is robust to
difficult occlusions both by other people and objects. Our method operates in
subsequent stages. The first stage is a convolutional neural network (CNN) that
estimates 2D and 3D pose features along with identity assignments for all
visible joints of all individuals. We contribute a new architecture for this
CNN, called SelecSLS Net, that uses novel selective long and short range skip
connections to improve the information flow allowing for a drastically faster
network without compromising accuracy. In the second stage, a fully-connected
neural network turns the possibly partial (on account of occlusion) 2D pose and
3D pose features for each subject into a complete 3D pose estimate per
individual. The third stage applies space-time skeletal model fitting to the
predicted 2D and 3D pose per subject to further reconcile the 2D and 3D pose,
and enforce temporal coherence. Our method returns the full skeletal pose in
joint angles for each subject. This is a further key distinction from previous
work that neither extracted global body positions nor joint angle results of a
coherent skeleton in real time for multi-person scenes. The proposed system
runs on consumer hardware at a previously unseen speed of more than 30 fps
given 512x320 images as input while achieving state-of-the-art accuracy, which
we will demonstrate on a range of challenging real-world scenes.

Modern approaches to pose and body shape estimation have recently achieved
strong performance even under challenging real-world conditions. Even from a
single image of a clothed person, a realistic looking body shape can be
inferred that captures a users' weight group and body shape type well. This
opens up a whole spectrum of applications -- in particular in fashion -- where
virtual try-on and recommendation systems can make use of these new and
automatized cues. However, a realistic depiction of the undressed body is
regarded highly private and therefore might not be consented by most people.
Hence, we ask if the automatic extraction of such information can be
effectively evaded. While adversarial perturbations have been shown to be
effective for manipulating the output of machine learning models -- in
particular, end-to-end deep learning approaches -- state of the art shape
estimation methods are composed of multiple stages. We perform the first
investigation of different strategies that can be used to effectively
manipulate the automatic shape estimation while preserving the overall
appearance of the original image.

Visual search is an important task, and it is part of daily human life. Thus, it has been a long-standing goal in Computer Vision to develop methods aiming at analysing human search intent and preferences. As the target of the search only exists in mind of the person, search intent prediction remains challenging for machine perception. In this thesis, we focus on advancing techniques for search target and preference prediction from implicit human cues. First, we propose a search target inference algorithm from human fixation data recorded during visual search. In contrast to previous work that has focused on individual instances as a search target in a closed world, we propose the first approach to predict the search target in open-world settings by learning the compatibility between observed fixations and potential search targets. Second, we further broaden the scope of search target prediction to categorical classes, such as object categories and attributes. However, state of the art models for categorical recognition, in general, require large amounts of training data, which is prohibitive for gaze data. To address this challenge, we propose a novel Gaze Pooling Layer that integrates gaze information into CNN-based architectures as an attention mechanism – incorporating both spatial and temporal aspects of human gaze behaviour. Third, we go one step further and investigate the feasibility of combining our gaze embedding approach, with the power of generative image models to visually decode, i.e. create a visual representation of, the search target. Forth, for the first time, we studied the effect of body shape on people preferences of outfits. We propose a novel and robust multi-photo approach to estimate the body shapes of each user and build a conditional model of clothing categories given body-shape. We demonstrate that in real-world data, clothing categories and body-shapes are correlated. We show that our approach estimates a realistic looking body shape that captures a user’s weight group and body shape type, even from a single image of a clothed person. However, an accurate depiction of the naked body is considered highly private and therefore, might not be consented by most people. First, we studied the perception of such technology via a user study. Then, in the last part of this thesis, we ask if the automatic extraction of such information can be effectively evaded. In summary, this thesis addresses several different tasks that aims to enable the vision system to analyse human search intent and preferences in real-world scenarios. In particular, the thesis proposes several novel ideas and models in visual search target prediction from human fixation data, for the first time studied the correlation between shape and clothing categories opening a new direction in clothing recommendation systems, and introduces a new topic in privacy and computer vision, aimed at preventing automatic 3D shape extraction from images.

Eye tracking and gaze-based human-computer interfaces have become a practical modality in desktop settings, since remote eye tracking is efficient and affordable. However, remote eye tracking remains constrained to indoor, laboratory-like conditions, in which lighting and user position need to be controlled. Mobile eye tracking has the potential to overcome these limitations and to allow people to move around freely and to use eye tracking on a daily basis during their everyday routine. However, mobile eye tracking currently faces two fundamental challenges that prevent it from being practically usable and that, consequently, have to be addressed before mobile eye tracking can truly be used by everyone: Mobile eye tracking needs to be advanced and made fully functional in unconstrained environments, and it needs to be made socially acceptable. Numerous sensing and analysis methods were initially developed for remote eye tracking and have been successfully applied for decades. Unfortunately, these methods are limited in terms of functionality and correctness, or even unsuitable for application in mobile eye tracking. Therefore, the majority of fundamental definitions, eye tracking methods, and gaze estimation approaches cannot be borrowed from remote eye tracking without adaptation. For example, the definitions of specific eye movements, like classical fixations, need to be extended to mobile settings where natural user and head motion are omnipresent. Corresponding analytical methods need to be adjusted or completely reimplemented based on novel approaches encoding the human gaze behaviour. Apart from these technical challenges, an entirely new, and yet under-explored, topic required for the breakthrough of mobile eye tracking as everyday technology is the overcoming of social obstacles. A first crucial key issue to defuse social objections is the building of acceptance towards mobile eye tracking. Hence, it is essential to replace the bulky appearance of current head-mounted eye trackers with an unobtrusive, appealing, and trendy design. The second high-priority theme of increasing importance for everyone is privacy and its protection, given that research and industry have not focused on or taken care of this problem at all. To establish true confidence, future devices have to find a fine balance between protecting users’ and bystanders’ privacy and attracting and convincing users of their necessity, utility, and potential with useful and beneficial features. The solution of technical challenges and social obstacles is the prerequisite for the development of a variety of novel and exciting applications in order to establish mobile eye tracking as a new paradigm, which ease our everyday life. This thesis addresses core technical challenges of mobile eye tracking that currently prevent it from being widely adopted. Specifically, this thesis proves that 3D data used for the calibration of mobile eye trackers improves gaze estimation and significantly reduces the parallax error. Further, it presents the first effective fixation detection method for head-mounted devices that is robust against the prevalence of user and gaze target motion. In order to achieve social acceptability, this thesis proposes an innovative and unobtrusive design for future mobile eye tracking devices and builds the first prototype with fully frame-embedded eye cameras combined with a calibration-free deep-trained appearance-based gaze estimation approach. To protect users’ and bystanders’ privacy in the presence of head-mounted eye trackers, this thesis presents another first-of-its-kind prototype. It is able to identify privacy-sensitive situations to automatically enable and disable the eye tracker’s first-person camera by means of a mechanical shutter, leveraging the combination of deep scene and eye movement features. Nevertheless, solving technical challenges and social obstacles alone is not sufficient to make mobile eye tracking attractive for the masses. The key to success is the development of convincingly useful, innovative, and essential applications. To extend the protection of users’ privacy on the software side as well, this thesis presents the first privacy-aware VR gaze interface using differential privacy. This method adds noise to recorded eye tracking data so that privacy-sensitive information like a user’s gender or identity is protected without impeding the utility of the data itself. In addition, the first large-scale online survey is conducted to understand users’ concerns with eye tracking. To develop and evaluate novel applications, this thesis presents the first publicly available long-term eye tracking datasets. They are used to show the unsupervised detection of users’ activities from eye movements alone using novel and efficient video-based encoding approaches as well as to propose the first proof-of-concept method to forecast users’ attentive behaviour during everyday mobile interactions from phone-integrated and body-worn sensors. This opens up possibilities for the development of a variety of novel and exciting applications. With more advanced features, accompanied by technological progress and sensor miniaturisation, eye tracking is increasingly integrated into conventional glasses as well as virtual and augmented reality (VR/AR) head-mounted displays, becoming an integral component of mobile interfaces. This thesis paves the way for the development of socially acceptable, privacy-aware, but highly functional mobile eye tracking devices and novel applications, so that mobile eye tracking can develop its full potential to become an everyday technology for everyone.

Adversarial training is the standard to train models robust against
adversarial examples. However, especially for complex datasets, adversarial
training incurs a significant loss in accuracy and is known to generalize
poorly to stronger attacks, e.g., larger perturbations or other threat models.
In this paper, we introduce confidence-calibrated adversarial training (CCAT)
where the key idea is to enforce that the confidence on adversarial examples
decays with their distance to the attacked examples. We show that CCAT
preserves better the accuracy of normal training while robustness against
adversarial examples is achieved via confidence thresholding, i.e., detecting
adversarial examples based on their confidence. Most importantly, in strong
contrast to adversarial training, the robustness of CCAT generalizes to larger
perturbations and other threat models, not encountered during training. For
evaluation, we extend the commonly used robust test error to our detection
setting, present an adaptive attack with backtracking and allow the attacker to
select, per test example, the worst-case adversarial example from multiple
black- and white-box attacks. We present experimental results using $L_\infty$,
$L_2$, $L_1$ and $L_0$ attacks on MNIST, SVHN and Cifar10.

While great progress has been made recently in automatic image manipulation,
it has been limited to object centric images like faces or structured scene
datasets. In this work, we take a step towards general scene-level image
editing by developing an automatic interaction-free object removal model. Our
model learns to find and remove objects from general scene images using
image-level labels and unpaired data in a generative adversarial network (GAN)
framework. We achieve this with two key contributions: a two-stage editor
architecture consisting of a mask generator and image in-painter that
co-operate to remove objects, and a novel GAN based prior for the mask
generator that allows us to flexibly incorporate knowledge about object shapes.
We experimentally show on two datasets that our method effectively removes a
wide variety of objects using weak supervision only

Deep neural perception and control networks have become key com-
ponents of self-driving vehicles. User acceptance is likely to benefit from easy-
to-interpret textual explanations which allow end-users to understand what trig-
gered a particular behavior. Explanations may be triggered by the neural con-
troller, namely
introspective explanations
, or informed by the neural controller’s
output, namely
rationalizations
. We propose a new approach to introspective ex-
planations which consists of two parts. First, we use a visual (spatial) attention
model to train a convolutional network end-to-end from images to the vehicle
control commands,
i
.
e
., acceleration and change of course. The controller’s at-
tention identifies image regions that potentially influence the network’s output.
Second, we use an attention-based video-to-text model to produce textual ex-
planations of model actions. The attention maps of controller and explanation
model are aligned so that explanations are grounded in the parts of the scene that
mattered to the controller. We explore two approaches to attention alignment,
strong- and weak-alignment. Finally, we explore a version of our model that
generates rationalizations, and compare with introspective explanations on the
same video segments. We evaluate these models on a novel driving dataset with
ground-truth human explanations, the Berkeley DeepDrive eXplanation (BDD-
X) dataset. Code is available at
github.com/JinkyuKimUCB/explainable-deep-driving

Encouraged by the recent progress in pedestrian detection, we investigate the gap between current state-of-the-art methods

and the “perfect single frame detector”. We enable our analysis by creating a human baseline for pedestrian detection (over the Caltech

pedestrian dataset). After manually clustering the frequent errors of a top detector, we characterise both localisation and background-

versus-foreground errors.

To address localisation errors we study the impact of training annotation noise on the detector performance, and show that we can

improve results even with a small portion of sanitised training data. To address background/foreground discrimination, we study convnets

for pedestrian detection, and discuss which factors affect their performance.

Other than our in-depth analysis, we report top performance on the Caltech pedestrian dataset, and provide a new sanitised set of

training and test annotations.

Natural language explanations of deep neural network decisions provide an
intuitive way for a AI agent to articulate a reasoning process. Current textual
explanations learn to discuss class discriminative features in an image.
However, it is also helpful to understand which attributes might change a
classification decision if present in an image (e.g., "This is not a Scarlet
Tanager because it does not have black wings.") We call such textual
explanations counterfactual explanations, and propose an intuitive method to
generate counterfactual explanations by inspecting which evidence in an input
is missing, but might contribute to a different classification decision if
present in the image. To demonstrate our method we consider a fine-grained
image classification task in which we take as input an image and a
counterfactual class and output text which explains why the image does not
belong to a counterfactual class. We then analyze our generated counterfactual
explanations both qualitatively and quantitatively using proposed automatic
metrics.

The inner structure of a material is called microstructure. It stores the

genesis of a material and determines all its physical and chemical properties.

While microstructural characterization is widely spread and well known, the

microstructural classification is mostly done manually by human experts, which

opens doors for huge uncertainties. Since the microstructure could be a

combination of different phases with complex substructures its automatic

classification is very challenging and just a little work in this field has

been carried out. Prior related works apply mostly designed and engineered

features by experts and classify microstructure separately from feature

extraction step. Recently Deep Learning methods have shown surprisingly good

performance in vision applications by learning the features from data together

with the classification step. In this work, we propose a deep learning method

for microstructure classification in the examples of certain microstructural

constituents of low carbon steel. This novel method employs pixel-wise

segmentation via Fully Convolutional Neural Networks (FCNN) accompanied by

max-voting scheme. Our system achieves 93.94% classification accuracy,

drastically outperforming the state-of-the-art method of 48.89% accuracy,

indicating the effectiveness of pixel-wise approaches. Beyond the success

presented in this paper, this line of research offers a more robust and first

of all objective way for the difficult task of steel quality appreciation.

The matching of multiple objects (e.g. shapes or images) is a fundamental
problem in vision and graphics. In order to robustly handle ambiguities, noise
and repetitive patterns in challenging real-world settings, it is essential to
take geometric consistency between points into account. Computationally, the
multi-matching problem is difficult. It can be phrased as simultaneously
solving multiple (NP-hard) quadratic assignment problems (QAPs) that are
coupled via cycle-consistency constraints. The main limitations of existing
multi-matching methods are that they either ignore geometric consistency and
thus have limited robustness, or they are restricted to small-scale problems
due to their (relatively) high computational cost. We address these
shortcomings by introducing a Higher-order Projected Power Iteration method,
which is (i) efficient and scales to tens of thousands of points, (ii)
straightforward to implement, (iii) able to incorporate geometric consistency,
and (iv) guarantees cycle-consistent multi-matchings. Experimentally we show
that our approach is superior to existing methods.

For autonomous agents to successfully operate in the real world, anticipation

of future events and states of their environment is a key competence. This

problem can be formalized as a sequence prediction problem, where a number of

observations are used to predict the sequence into the future. However,

real-world scenarios demand a model of uncertainty of such predictions, as

future states become increasingly uncertain and multi-modal -- in particular on

long time horizons. This makes modelling and learning challenging. We cast

state of the art semantic segmentation and future prediction models based on

deep learning into a Bayesian formulation that in turn allows for a full

Bayesian treatment of the prediction problem. We present a new sampling scheme

for this model that draws from the success of variational autoencoders by

incorporating a recognition network. In the experiments we show that our model

outperforms prior work in accuracy of the predicted segmentation and provides

calibrated probabilities that also better capture the multi-modal aspects of

possible future states of street scenes.

We introduce Primal-Dual Wasserstein GAN, a new learning algorithm for

building latent variable models of the data distribution based on the primal

and the dual formulations of the optimal transport (OT) problem. We utilize the

primal formulation to learn a flexible inference mechanism and to create an

optimal approximate coupling between the data distribution and the generative

model. In order to learn the generative model, we use the dual formulation and

train the decoder adversarially through a critic network that is regularized by

the approximate coupling obtained from the primal. Unlike previous methods that

violate various properties of the optimal critic, we regularize the norm and

the direction of the gradients of the critic function. Our model shares many of

the desirable properties of auto-encoding models in terms of mode coverage and

latent structure, while avoiding their undesirable averaging properties, e.g.

their inability to capture sharp visual features when modeling real images. We

compare our algorithm with several other generative modeling techniques that

utilize Wasserstein distances on Frechet Inception Distance (FID) and Inception

Scores (IS).

With the widespread use of machine learning (ML) techniques, ML as a service
has become increasingly popular. In this setting, an ML model resides on a
server and users can query the model with their data via an API. However, if
the user's input is sensitive, sending it to the server is not an option.
Equally, the service provider does not want to share the model by sending it to
the client for protecting its intellectual property and pay-per-query business
model. In this paper, we propose MLCapsule, a guarded offline deployment of
machine learning as a service. MLCapsule executes the machine learning model
locally on the user's client and therefore the data never leaves the client.
Meanwhile, MLCapsule offers the service provider the same level of control and
security of its model as the commonly used server-side execution. In addition,
MLCapsule is applicable to offline applications that require local execution.
Beyond protecting against direct model access, we demonstrate that MLCapsule
allows for implementing defenses against advanced attacks on machine learning
models such as model stealing/reverse engineering and membership inference.

In this study, we explore building a two-stage framework for enabling users

to directly manipulate high-level attributes of a natural scene. The key to our

approach is a deep generative network which can hallucinate images of a scene

as if they were taken at a different season (e.g. during winter), weather

condition (e.g. in a cloudy day) or time of the day (e.g. at sunset). Once the

scene is hallucinated with the given attributes, the corresponding look is then

transferred to the input image while preserving the semantic details intact,

giving a photo-realistic manipulation result. As the proposed framework

hallucinates what the scene will look like, it does not require any reference

style image as commonly utilized in most of the appearance or style transfer

approaches. Moreover, it allows to simultaneously manipulate a given scene

according to a diverse set of transient attributes within a single model,

eliminating the need of training multiple networks per each translation task.

Our comprehensive set of qualitative and quantitative results demonstrate the

effectiveness of our approach against the competing methods.

We propose a novel approach to jointly perform 3D object retrieval and pose
estimation from monocular images.In order to make the method robust to real
world scene variations in the images, e.g. texture, lighting and background,we
learn an embedding space from 3D data that only includes the relevant
information, namely the shape and pose.Our method can then be trained for
robustness under real world scene variations without having to render a large
training set simulating these variations. Our learned embedding explicitly
disentangles a shape vector and a pose vector, which alleviates both pose bias
for 3D shape retrieval and categorical bias for pose estimation. Having the
learned disentangled embedding, we train a CNN to map the images to the
embedding space, and then retrieve the closest 3D shape from the database and
estimate the 6D pose of the object using the embedding vectors. Our method
achieves 10.8 median error for pose estimation and 0.514 top-1-accuracy for
category agnostic 3D object retrieval on the Pascal3D+ dataset. It therefore
outperforms the previous state-of-the-art methods on both tasks.

From autonomous driving cars to surgical robots, robotic system has enjoyed significant growth over the past decade. With the rapid development in robotics alongside the evolution in the related fields, such as computer vision and machine learning, integrating perception, anticipation and manipulation is key to the success of future robotic system. In this thesis, we explore different ways of such integration to extend the capabilities of a robotic system to take on more challenging real world tasks. On anticipation and perception, we address the recognition of ongoing activity from videos. In particular we focus on long-duration and complex activities and hence propose a new challenging dataset to facilitate the work. We introduce hierarchical labels over the activity classes and investigate the temporal accuracy-specificity trade-offs. We propose a new method based on recurrent neural networks that learns to predict over this hierarchy and realize accuracy specificity trade-offs. Our method outperforms several baselines on this new challenge. On manipulation with perception, we propose an efficient framework for programming a robot to use human tools. We first present a novel and compact model for using tools described by a tip model. Then we explore a strategy of utilizing a dual-gripper approach for manipulating tools – motivated by the absence of dexterous hands on widely available general purpose robots. Afterwards, we embed the tool use learning into a hierarchical architecture and evaluate it on a Baxter research robot. Finally, combining perception, anticipation and manipulation, we focus on a block stacking task. First we explore how to guide robot to place a single block into the scene without collapsing the existing structure. We introduce a mechanism to predict physical stability directly from visual input and evaluate it first on a synthetic data and then on real-world block stacking. Further, we introduce the target stacking task where the agent stacks blocks to reproduce a tower shown in an image. To do so, we create a synthetic block stacking environment with physics simulation in which the agent can learn block stacking end-to-end through trial and error, bypassing to explicitly model the corresponding physics knowledge. We propose a goal-parametrized GDQN model to plan with respect to the specific goal. We validate the model on both a navigation task in a classic gridworld environment and the block stacking task.

We propose a deep representation of appearance, i. e. the relation of color,

surface orientation, viewer position, material and illumination. Previous

approaches have used deep learning to extract classic appearance

representations relating to reflectance model parameters (e. g. Phong) or

illumination (e. g. HDR environment maps). We suggest to directly represent

appearance itself as a network we call a deep appearance map (DAM). This is a

4D generalization over 2D reflectance maps, which held the view direction

fixed. First, we show how a DAM can be learned from images or video frames and

later be used to synthesize appearance, given new surface orientations and

viewer positions. Second, we demonstrate how another network can be used to map

from an image or video frames to a DAM network to reproduce this appearance,

without using a lengthy optimization such as stochastic gradient descent

(learning-to-learn). Finally, we generalize this to an appearance

estimation-and-segmentation task, where we map from an image showing multiple

materials to multiple networks reproducing their appearance, as well as

per-pixel segmentation.

Machine learning is transforming the world. Its application areas span privacy
sensitive and security critical tasks such as human identification and self-driving
cars. These applications raise privacy and security related questions that are not
fully understood or answered yet: Can automatic person recognisers identify people
in photos even when their faces are blurred? How easy is it to find an adversarial
input for a self-driving car that makes it drive off the road?
This thesis contributes one of the first steps towards a better understanding of
such concerns. We observe that many privacy and security critical scenarios for
learned models involve input data manipulation: users obfuscate their identity by
blurring their faces and adversaries inject imperceptible perturbations to the input
signal. We introduce a data manipulator framework as a tool for collectively describing
and analysing privacy and security relevant scenarios involving learned models.
A data manipulator introduces a shift in data distribution for achieving privacy or
security related goals, and feeds the transformed input to the target model. This
framework provides a common perspective on the studies presented in the thesis.
We begin the studies from the user’s privacy point of view. We analyse the
efficacy of common obfuscation methods like face blurring, and show that they
are surprisingly ineffective against state of the art person recognition systems. We
then propose alternatives based on head inpainting and adversarial examples. By
studying the user privacy, we also study the dual problem: model security. In model
security perspective, a model ought to be robust and reliable against small amounts
of data manipulation. In both cases, data are manipulated with the goal of changing
the target model prediction. User privacy and model security problems can be
described with the same objective.
We then study the knowledge aspect of the data manipulation problem. The more
one knows about the target model, the more effective manipulations one can craft.
We propose a game theoretic manipulation framework to systematically represent
the knowledge level on the target model and derive privacy and security guarantees.
We then discuss ways to increase knowledge about a black-box model by only querying
it, deriving implications that are relevant to both privacy and security perspectives.

Machine Learning techniques are widely used by online services (e.g. Google,

Apple) in order to analyze and make predictions on user data. As many of the

provided services are user-centric (e.g. personal photo collections, speech

recognition, personal assistance), user data generated on personal devices is

key to provide the service. In order to protect the data and the privacy of the

user, federated learning techniques have been proposed where the data never

leaves the user's device and "only" model updates are communicated back to the

server. In our work, we propose a new threat model that is not concerned with

learning about the content - but rather is concerned with the linkability of

users during such decentralized learning scenarios.

We show that model updates are characteristic for users and therefore lend

themselves to linkability attacks. We show identification and matching of users

across devices in closed and open world scenarios. In our experiments, we find

our attacks to be highly effective, achieving 20x-175x chance-level

performance.

In order to mitigate the risks of linkability attacks, we study various

strategies. As adding random noise does not offer convincing operation points,

we propose strategies based on using calibrated domain-specific data; we find

these strategies offers substantial protection against linkability threats with

little effect to utility.

We propose a novel end-to-end trainable framework for the graph decomposition
problem. The minimum cost multicut problem is first converted to an
unconstrained binary cubic formulation where cycle consistency constraints are
incorporated into the objective function. The new optimization problem can be
viewed as a Conditional Random Field (CRF) in which the random variables are
associated with the binary edge labels of the initial graph and the hard
constraints are introduced in the CRF as high-order potentials. The parameters
of a standard Neural Network and the fully differentiable CRF are optimized in
an end-to-end manner. Furthermore, our method utilizes the cycle constraints as
meta-supervisory signals during the learning of the deep feature
representations by taking the dependencies between the output random variables
into account. We present analyses of the end-to-end learned representations,
showing the impact of the joint training, on the task of clustering images of
MNIST. We also validate the effectiveness of our approach both for the feature
learning and the final clustering on the challenging task of real-world
multi-person pose estimation.

As first-person cameras in head-mounted displays become increasingly

prevalent, so does the problem of infringing user and bystander privacy. To

address this challenge, we present PrivacEye, a proof-of-concept system that

detects privacysensitive everyday situations and automatically enables and

disables the first-person camera using a mechanical shutter. To close the

shutter, PrivacEye detects sensitive situations from first-person camera videos

using an end-to-end deep-learning model. To open the shutter without visual

input, PrivacEye uses a separate, smaller eye camera to detect changes in

users' eye movements to gauge changes in the "privacy level" of the current

situation. We evaluate PrivacEye on a dataset of first-person videos recorded

in the daily life of 17 participants that they annotated with privacy

sensitivity levels. We discuss the strengths and weaknesses of our

proof-of-concept system based on a quantitative technical evaluation as well as

qualitative insights from semi-structured interviews.

Following a period of expedited progress in the capabilities of digital systems, the society begins to realize that systems designed to assist people in various tasks can also harm individuals and society. Mediating access to information and explicitly or implicitly ranking people in increasingly many applications, search systems have a substantial potential to contribute to such unwanted outcomes. Since they collect vast amounts of data about both searchers and search subjects, they have the potential to violate the privacy of both of these groups of users. Moreover, in applications where rankings influence people's economic livelihood outside of the platform, such as sharing economy or hiring support websites, search engines have an immense economic power over their users in that they control user exposure in ranked results. This thesis develops new models and methods broadly covering different aspects of privacy and fairness in search systems for both searchers and search subjects. Specifically, it makes the following contributions: (1) We propose a model for computing individually fair rankings where search subjects get exposure proportional to their relevance. The exposure is amortized over time using constrained optimization to overcome searcher attention biases while preserving ranking utility. (2) We propose a model for computing sensitive search exposure where each subject gets to know the sensitive queries that lead to her profile in the top-k search results. The problem of finding exposing queries is technically modeled as reverse nearest neighbor search, followed by a weekly-supervised learning to rank model ordering the queries by privacy-sensitivity. (3) We propose a model for quantifying privacy risks from textual data in online communities. The method builds on a topic model where each topic is annotated by a crowdsourced sensitivity score, and privacy risks are associated with a user's relevance to sensitive topics. We propose relevance measures capturing different dimensions of user interest in a topic and show how they correlate with human risk perceptions. (4) We propose a model for privacy-preserving personalized search where search queries of different users are split and merged into synthetic profiles. The model mediates the privacy-utility trade-off by keeping semantically coherent fragments of search histories within individual profiles, while trying to minimize the similarity of any of the synthetic profiles to the original user profiles. The models are evaluated using information retrieval techniques and user studies over a variety of datasets, ranging from query logs, through social media and community question answering postings, to item listings from sharing economy platforms.