no code implementations • 18 Apr 2024 • Christoph Reich, Oliver Hahn, Daniel Cremers, Stefan Roth, Biplob Debnath
The use of standardized codecs, such as JPEG or H. 264, is prevalent and required to ensure interoperability.
no code implementations • 26 Jan 2024 • Chu Li, Stefan Roth, Aydin Sezgin
Simulations verify the correctness of the closed-form expressions and demonstrate that we can effectively improve the secrecy rate, especially when the eavesdropper is close to the RIS or the legitimate user.
no code implementations • 11 Oct 2023 • Sascha Rosbach, Stefan M. Leupold, Simon Großjohann, Stefan Roth
Automated vehicles operating in urban environments have to reliably interact with other traffic participants.
1 code implementation • ICCV 2023 • Gopika Sudhakaran, Devendra Singh Dhami, Kristian Kersting, Stefan Roth
Recent years have seen a growing interest in Scene Graph Generation (SGG), a comprehensive visual scene understanding task that aims to predict entity relationships using a relation encoder-decoder pipeline stacked on top of an object encoder-decoder backbone.
1 code implementation • ICCV 2023 • Robin Hesse, Simone Schaub-Meyer, Stefan Roth
Using our tools, we report results for 24 different combinations of neural models and XAI methods, demonstrating the strengths and weaknesses of the assessed methods in a fully automatic and systematic manner.
Explainable artificial intelligence Explainable Artificial Intelligence (XAI)
1 code implementation • 16 May 2023 • Robin Hesse, Simone Schaub-Meyer, Stefan Roth
Many convolutional neural networks (CNNs) rely on progressive downsampling of their feature maps to increase the network's receptive field and decrease computational cost.
1 code implementation • 25 Nov 2022 • Moritz Nottebaum, Stefan Roth, Simone Schaub-Meyer
The feature extraction layers help to compress the input and extract relevant information for the latter stages, such as motion estimation.
1 code implementation • 22 Nov 2022 • Krishnakant Singh, Simone Schaub-Meyer, Stefan Roth
In addition, methods that use semantic masks to edit images have difficulty preserving the identity and are unable to perform controlled style edits.
1 code implementation • 10 Aug 2022 • Sherwin Bahmani, Oliver Hahn, Eduard Zamfir, Nikita Araslanov, Daniel Cremers, Stefan Roth
The lack of out-of-domain generalization is a critical weakness of deep networks for semantic segmentation.
1 code implementation • 3 May 2022 • Franz Klein, Shweta Mahajan, Stefan Roth
Stylized image captioning as presented in prior work aims to generate captions that reflect characteristics beyond a factual description of the scene composition, such as sentiments.
1 code implementation • NeurIPS 2021 • Robin Hesse, Simone Schaub-Meyer, Stefan Roth
Mitigating the dependence on spurious correlations present in the training dataset is a quickly emerging and important topic of deep learning.
1 code implementation • NeurIPS 2021 • Nikita Araslanov, Simone Schaub-Meyer, Stefan Roth
On established VOS benchmarks, our approach exceeds the segmentation accuracy of previous work despite using significantly less training data and compute power.
1 code implementation • ICCV 2021 • Shweta Mahajan, Stefan Roth
Autoregressive models are a class of exact inference approaches with highly flexible functional forms, yielding state-of-the-art density estimates for natural images.
no code implementations • 29 Sep 2021 • Sherwin Bahmani, Oliver Hahn, Eduard Sebastian Zamfir, Nikita Araslanov, Stefan Roth
In this work, we empirically study an adaptive inference strategy for semantic segmentation that adjusts the model to the test sample before producing the final prediction.
no code implementations • 29 Sep 2021 • Jannik Schmitt, Stefan Roth
While deep neural networks for classification have shown impressive predictive performance, e. g. in image classification, they generally tend to be overconfident.
1 code implementation • Findings (ACL) 2022 • Jonas Pfeiffer, Gregor Geigle, Aishwarya Kamath, Jan-Martin O. Steitz, Stefan Roth, Ivan Vulić, Iryna Gurevych
In this work, we address this gap and provide xGQA, a new multilingual evaluation benchmark for the visual question answering task.
no code implementations • 9 Sep 2021 • Jan-Martin O. Steitz, Jonas Pfeiffer, Iryna Gurevych, Stefan Roth
Reasoning over multiple modalities, e. g. in Visual Question Answering (VQA), requires an alignment of semantic concepts across domains.
no code implementations • CVPR 2021 • Jiangxin Dong, Stefan Roth, Bernt Schiele
The classical maximum a-posteriori (MAP) framework for non-blind image deblurring requires defining suitable data and regularization terms, whose interplay yields the desired clear image through optimization.
1 code implementation • CVPR 2021 • Junhwa Hur, Stefan Roth
Estimating 3D scene flow from a sequence of monocular images has been gaining increased attention due to the simple, economical capture setup.
1 code implementation • CVPR 2021 • Nikita Araslanov, Stefan Roth
We propose an approach to domain adaptation for semantic segmentation that is both practical and highly accurate.
Ranked #16 on Domain Adaptation on SYNTHIA-to-Cityscapes
1 code implementation • NeurIPS 2020 • Jiangxin Dong, Stefan Roth, Bernt Schiele
We present a simple and effective approach for non-blind image deblurring, combining classical techniques and deep learning.
no code implementations • 15 Mar 2021 • Jannik Schmitt, Stefan Roth
Variational inference methods for BNNs approximate the intractable weight posterior with a tractable distribution, yet mostly rely on sampling from the variational distribution during training and inference.
no code implementations • 12 Feb 2021 • Alexander Deisting, Abigail Victoria Waldron, Edward Atkin, Gary Barker, Anastasia Basharina-Freshville, Christopher Betancourt, Steven Boyd, Dominic Brailsford, Zachary Chen-Wishart, Linda Cremonesi, Adriana Dias, Patrick Dunne, Jennifer Haigh, Philip Hamacher-Baumann, Sebastian Jones, Asher Kaboth, Alexander Korzenev, William Ma, Philippe Mermod, Maria Mironova, Jocelyn Monroe, Ryan Nichol, Toby Nonnenmacher, Jaroslaw Nowak, William Parker, Harrison Ritchie-Yates, Stefan Roth, Ruben Saakyan, Nicola Serra, Yuri Shitov, Jochen Steinmann, Adam Tarrant, Melissa Uchida, Sammy Valder, Mark Ward, Morgan Wascko
Measurements of proton-nucleus scattering and high resolution neutrino-nucleus interaction imaging are key to reduce neutrino oscillation systematic uncertainties in future experiments.
Instrumentation and Detectors High Energy Physics - Experiment
no code implementations • 14 Dec 2020 • Stefan Roth, Stefano Tomasin, Marco Maso, Aydin Sezgin
We analyze the attack and assess the obtained localization accuracy against various parameters.
Information Theory Cryptography and Security Signal Processing Information Theory
1 code implementation • NeurIPS 2020 • Shweta Mahajan, Stefan Roth
Our framework not only enables diverse captioning through context-based pseudo supervision, but extends this to images with novel objects and without paired captions in the training data.
no code implementations • 15 Oct 2020 • Patrick Dendorfer, Aljoša Ošep, Anton Milan, Konrad Schindler, Daniel Cremers, Ian Reid, Stefan Roth, Laura Leal-Taixé
We present MOTChallenge, a benchmark for single-camera Multiple Object Tracking (MOT) launched in late 2014, to collect existing and new data, and create a framework for the standardized evaluation of multiple object tracking methods.
no code implementations • 11 Jul 2020 • Sascha Rosbach, Xing Li, Simon Großjohann, Silviu Homoceanu, Stefan Roth
Furthermore, the temporal attention mechanism learns persistent interaction with other vehicles over an extended planning horizon.
no code implementations • 28 May 2020 • Wentong Liao, Xiang Chen, Jingfeng Yang, Stefan Roth, Michael Goesele, Michael Ying Yang, Bodo Rosenhahn
This strengthens the local feature invariance for the resampled features and enables detecting vehicles in an arbitrary orientation.
1 code implementation • CVPR 2020 • Nikita Araslanov, Stefan Roth
This is in contrast to earlier work that used only a single stage $-$ training one segmentation network on image labels $-$ which was abandoned due to inferior segmentation accuracy.
1 code implementation • CVPR 2020 • Junhwa Hur, Stefan Roth
Our model achieves state-of-the-art accuracy among unsupervised/self-supervised learning approaches to monocular scene flow, and yields competitive results for the optical flow and monocular depth estimation sub-tasks.
1 code implementation • CVPR 2020 • Shweta Mahajan, Apratim Bhattacharyya, Mario Fritz, Bernt Schiele, Stefan Roth
Flow-based generative models are an important class of exact inference models that admit efficient inference and sampling for image synthesis.
no code implementations • 6 Apr 2020 • Junhwa Hur, Stefan Roth
Akin to many subareas of computer vision, the recent advances in deep learning have also significantly influenced the literature on optical flow.
1 code implementation • CVPR 2020 • Anne S. Wannenwetsch, Stefan Roth
Encoder-decoder networks have found widespread use in various dense prediction tasks.
1 code implementation • 19 Mar 2020 • Patrick Dendorfer, Hamid Rezatofighi, Anton Milan, Javen Shi, Daniel Cremers, Ian Reid, Stefan Roth, Konrad Schindler, Laura Leal-Taixé
The benchmark for Multiple Object Tracking, MOTChallenge, was launched with the goal to establish a standardized evaluation of multiple object tracking methods.
Multi-Object Tracking Multiple Object Tracking with Transformer +2
1 code implementation • ICLR 2020 • Shweta Mahajan, Iryna Gurevych, Stefan Roth
Therefore, we propose a novel semi-supervised framework, which models shared information between domains and domain-specific information separately.
no code implementations • 7 Dec 2019 • Sascha Rosbach, Vinit James, Simon Großjohann, Silviu Homoceanu, Xing Li, Stefan Roth
In this work, we propose a deep learning approach based on inverse reinforcement learning that generates situation-dependent reward functions.
1 code implementation • 26 Sep 2019 • Jochen Gast, Stefan Roth
In contrast to these involved models, we found that a simple baseline CNN can perform astonishingly well when particular care is taken w. r. t.
no code implementations • 26 Sep 2019 • Vladyslav Yushchenko, Nikita Araslanov, Stefan Roth
We identify two pathological cases of temporal inconsistencies in video generation: video freezing and video looping.
no code implementations • 14 Sep 2019 • Shweta Mahajan, Teresa Botschen, Iryna Gurevych, Stefan Roth
One of the key challenges in learning joint embeddings of multiple modalities, e. g. of images and text, is to ensure coherent cross-modal semantics that generalize across datasets.
1 code implementation • 9 Sep 2019 • Anne S. Wannenwetsch, Martin Kiefel, Peter V. Gehler, Stefan Roth
When adding our network layer to state-of-the-art networks for optical flow and semantic segmentation, boundary artifacts are removed and the accuracy is improved.
no code implementations • 10 Jun 2019 • Patrick Dendorfer, Hamid Rezatofighi, Anton Milan, Javen Shi, Daniel Cremers, Ian Reid, Stefan Roth, Konrad Schindler, Laura Leal-Taixe
Standardized benchmarks are crucial for the majority of computer vision applications.
no code implementations • 1 May 2019 • Sascha Rosbach, Vinit James, Simon Großjohann, Silviu Homoceanu, Stefan Roth
Behavior and motion planning play an important role in automated driving.
1 code implementation • CVPR 2019 • Nikita Araslanov, Constantin Rothkopf, Stefan Roth
Most approaches to visual scene analysis have emphasised parallel processing of the image elements.
2 code implementations • CVPR 2019 • Junhwa Hur, Stefan Roth
While leading to more accurate results, the downside of this is an increased number of parameters.
Ranked #10 on Optical Flow Estimation on KITTI 2012
2 code implementations • NeurIPS 2018 • Tobias Plötz, Stefan Roth
To exploit our relaxation, we propose the neural nearest neighbors block (N3 block), a novel non-local processing layer that leverages the principle of self-similarity and can be used as building block in modern neural network architectures.
Ranked #1 on Grayscale Image Denoising on Set12 sigma70
no code implementations • 4 Oct 2018 • Jan-Martin O. Steitz, Faraz Saeedan, Stefan Roth
First, we introduce a novel multi-view pooling layer to perform a 3D aggregation of 2D CNN-features extracted from each view.
no code implementations • ECCV 2018 • Meiguang Jin, Stefan Roth, Paolo Favaro
We introduce a family of novel approaches to single-image blind deconvolution, ie , the problem of recovering a sharp image and a blur kernel from a single blurry input.
no code implementations • SEMEVAL 2018 • Hatem Mousselly-Sergieh, Teresa Botschen, Iryna Gurevych, Stefan Roth
Current methods for knowledge graph (KG) representation learning focus solely on the structure of the KG and do not exploit any kind of external information, such as visual and linguistic information corresponding to the KG entities.
no code implementations • NAACL 2018 • Teresa Botschen, Iryna Gurevych, Jan-Christoph Klie, Hatem Mousselly-Sergieh, Stefan Roth
Our analysis shows that for the German data, textual representations are still competitive with multimodal ones.
4 code implementations • CVPR 2018 • Jochen Gast, Stefan Roth
Even though probabilistic treatments of neural networks have a long history, they have not found widespread use in practice.
3 code implementations • CVPR 2018 • Stephan R. Richter, Stefan Roth
We scale this baseline to higher resolutions by proposing a memory-efficient shape encoding, which recursively decomposes a 3D shape into nested shape layers, similar to the pieces of a Matryoshka doll.
Ranked #6 on 3D Object Reconstruction on Data3D−R2N2
2 code implementations • CVPR 2018 • Faraz Saeedan, Nicolas Weber, Michael Goesele, Stefan Roth
This is commonly referred to as pooling, and is applied to reduce the number of parameters, improve invariance to certain distortions, and increase the receptive field size.
no code implementations • CVPR 2018 • Tobias Plötz, Anne S. Wannenwetsch, Stefan Roth
In this paper we propose stochastic variational inference with gradient linearization (SVIGL).
2 code implementations • 21 Nov 2017 • Simon Meister, Junhwa Hur, Stefan Roth
By optionally fine-tuning on the KITTI training data, our method achieves competitive optical flow accuracy on the KITTI 2012 and 2015 benchmarks, thus in addition enabling generic pre-training of supervised networks for datasets with limited amounts of ground truth.
no code implementations • ICCV 2017 • Anne S. Wannenwetsch, Margret Keuper, Stefan Roth
We overcome the artificial separation of optical flow and confidence estimation by introducing a method that jointly predicts optical flow and its underlying uncertainty.
no code implementations • ICCV 2017 • Junhwa Hur, Stefan Roth
The key feature of our model is to fully exploit the symmetry properties that characterize optical flow and occlusions in the two consecutive images.
no code implementations • CVPR 2017 • Tobias Plötz, Stefan Roth
Lacking realistic ground truth data, image denoising techniques are traditionally evaluated on images corrupted by synthesized i. i. d.
no code implementations • 5 Jul 2017 • Florian Lang, Tobias Plötz, Stefan Roth
While the concept is appealing, we show that the original reconstruction approach assumes noise-free measurements and quickly breaks down otherwise.
no code implementations • CVPR 2017 • Meiguang Jin, Stefan Roth, Paolo Favaro
We present a novel approach to noise-blind deblurring, the problem of deblurring an image with known blur, but unknown noise level.
no code implementations • 10 Apr 2017 • Laura Leal-Taixé, Anton Milan, Konrad Schindler, Daniel Cremers, Ian Reid, Stefan Roth
Standardized benchmarks are crucial for the majority of computer vision applications.
no code implementations • 2 Apr 2017 • Marius Cordts, Timo Rehfeld, Lukas Schneider, David Pfeiffer, Markus Enzweiler, Stefan Roth, Marc Pollefeys, Uwe Franke
We believe this challenge should be faced by introducing a representation of the sensory data that provides compressed and structured access to all relevant visual content of the scene.
2 code implementations • 7 Aug 2016 • Stephan R. Richter, Vibhav Vineet, Stefan Roth, Vladlen Koltun
Recent progress in computer vision has been driven by high-capacity models trained on large datasets.
no code implementations • 28 Jul 2016 • Anita Sellent, Carsten Rother, Stefan Roth
With this paper we are the first to show how the availability of stereo video can aid the challenging video deblurring task.
no code implementations • 26 Jul 2016 • Junhwa Hur, Stefan Roth
The importance and demands of visual scene understanding have been steadily increasing along with the active development of autonomous systems.
no code implementations • CVPR 2016 • Jochen Gast, Anita Sellent, Stefan Roth
A two-stage pipeline, first in derivative space and then in image space, allows to estimate both parametric object motion as well as a motion segmentation from a single image alone.
1 code implementation • CVPR 2016 • Marius Cordts, Mohamed Omran, Sebastian Ramos, Timo Rehfeld, Markus Enzweiler, Rodrigo Benenson, Uwe Franke, Stefan Roth, Bernt Schiele
Visual understanding of complex urban street scenes is an enabling factor for a wide range of applications.
8 code implementations • 2 Mar 2016 • Anton Milan, Laura Leal-Taixe, Ian Reid, Stefan Roth, Konrad Schindler
Recently, a new benchmark for Multiple Object Tracking, MOTChallenge, was launched with the goal of collecting existing and new data and creating a framework for the standardized evaluation of multiple object tracking methods.
no code implementations • ICCV 2015 • Tobias Plotz, Stefan Roth
Many existing approaches for image-to-geometry registration assume that either a textured 3D model or a good initial guess of the 3D pose is available to bootstrap the registration process.
no code implementations • CVPR 2015 • Stephan R. Richter, Stefan Roth
Von Mises-Fisher distributions in the leaves of each tree enable the estimation of surface normals.
2 code implementations • 8 Apr 2015 • Laura Leal-Taixé, Anton Milan, Ian Reid, Stefan Roth, Konrad Schindler
We discuss the challenges of creating such a framework, collecting existing and new data, gathering state-of-the-art methods to be tested on the datasets, and finally creating a unified evaluation system.
no code implementations • CVPR 2014 • Uwe Schmidt, Stefan Roth
To that end we propose shrinkage fields, a random field-based architecture that combines the image model and the optimization algorithm in a single unit.
no code implementations • 8 Apr 2014 • Uwe Schmidt, Jeremy Jancsary, Sebastian Nowozin, Stefan Roth, Carsten Rother
We posit two reasons for this: First, the blur kernel is often only known at test time, requiring any discriminative approach to cope with considerable variability.
no code implementations • CVPR 2013 • Uwe Schmidt, Carsten Rother, Sebastian Nowozin, Jeremy Jancsary, Stefan Roth
From this analysis, we derive a discriminative model cascade for image deblurring.
no code implementations • CVPR 2013 • Anton Milan, Konrad Schindler, Stefan Roth
When tracking multiple targets in crowded scenarios, modeling mutual exclusion between distinct targets becomes important at two levels: (1) in data association, each target observation should support at most one trajectory and each trajectory should be assigned at most one observation per frame; (2) in trajectory estimation, two trajectories should remain spatially separated at all times to avoid collisions.
1 code implementation • International Journal of Computer Vision 2010 • Simon Baker, Daniel Scharstein, J. P. Lewis, Stefan Roth, Michael J. Black, Richard Szeliski
The quantitative evaluation of optical flow algorithms by Barron et al. (1994) led to significant advances in performance.