Accepted Papers

Accepted Papers:

Best Paper:

Robust Semantic Segmentation by Redundant Networks With a Layer-Specific Loss Contribution and Majority Vote

Andreas Bär (Technische Universität Braunschweig)*; Marvin R Klingner (Technische Universität Braunschweig ); Serin Varghese (Volkswagen AG); Fabian Hüger (Volkswagen Group Research); Peter Schlicht (Volkswagen Group Research); Tim Fingscheidt ( Technische Universität Braunschweig)

Second Best Paper:

Mind the Gap - A Benchmark for Dense Depth Prediction beyond Lidar

Hendrik Schilling (rabbitAI)*; Marcel Gutsche (rabbitAI); Alexander Brock (Heidelberg University); Dane Späth (Heidelberg University); Carsten Rother (University of Heidelberg); Karsten Krispin (rabbitAI)

Orals:

Improved Noise and Attack Robustness for Semantic Segmentation by Using Multi-Task Training with Self-Supervised Depth Estimation

Marvin R Klingner (Technische Universität Braunschweig )*; Andreas Bär (Technische Universität Braunschweig); Tim Fingscheidt ( Technische Universität Braunschweig)

Attentional Bottleneck: Towards an Interpretable Deep Driving Network

Jinkyu Kim (UC Berkeley)*; Mayank Bansal (.)

Detection and Retrieval of Out-of-Distribution Objects in Semantic Segmentation

Philipp Oberdiek (TU Dortmund)*; Matthias Rottmann (University of Wuppertal); Gernot Fink (TU Dortmund)

Robust Semantic Segmentation by Redundant Networks With a Layer-Specific Loss Contribution and Majority Vote

Mind the Gap - A Benchmark for Dense Depth Prediction beyond Lidar

Explaining Autonomous Driving by Learning End-to-End Visual Attention

Luca Cultrera (University of Florence); Lorenzo Seidenari (University of Florence); Federico Becattini (Università di Firenze)*; Pietro Pala (University of Florence); Alberto Del Bimbo (University of Florence)

Posters:

Evaluating Scalable Bayesian Deep Learning Methods for Robust Computer Vision

Fredrik K Gustafsson (Uppsala University)*; Martin Danelljan (ETH Zurich); Thomas Schön (Uppsala University)

Leveraging combinatorial testing for safety-critical computer vision datasets

Christoph D Gladisch (Robert Bosch GmbH)*; Christian Heinzemann (Robert Bosch GmbH); Martin Herrmann (Robert Bosch GmbH); Matthias Woehrle (Robert Bosch GmbH)

Multivariate Confidence Calibration for Object Detection

Fabian Küppers (Ruhr West University of Applied Sciences)*; Jan Kronenberger (Ruhr West University of Applied Sciences); Amirhossein Shantia (Visteon Electronics GmbH); Anselm Haselhoff (Ruhr West University of Applied Sciences)

Generating Socially Acceptable Perturbations for Efficient Evaluation of Autonomous Vehicles

Songan Zhang (University of Michigan)*

Self-Supervised Domain Mismatch Estimation for Autonomous Perception

Jonas Löhdefink (Institute for Communications Technology)*; Justin Fehrling (Institute for Communications Technology); Marvin R Klingner (Technische Universität Braunschweig ); Fabian Hüger (Volkswagen Group Research); Peter Schlicht (Volkswagen Group Research); Nico M Schmidt (Volkswagen AG); Tim Fingscheidt ( Technische Universität Braunschweig)

Unsupervised Temporal Consistency Metric for Video Segmentation in Highly-Automated Driving

Serin Varghese (Volkswagen AG)*; Yasin Bayzidi (Volkswagen AG); Andreas Bär (Technische Universität Braunschweig); Nikhil Kapoor (Volkswagen AG); Sounak Lahiri (Volkswagen AG); Jan David Schneider (Volkswagen AG); Nico M Schmidt (Volkswagen AG); Peter Schlicht (Volkswagen Group Research); Fabian Hüger (Volkswagen Group Research); Tim Fingscheidt ( Technische Universität Braunschweig)

Using Mixture of Expert Models to Gain Insights into Semantic Segmentation

Svetlana Pavlitskaya (FZI Research Center for Information Technology)*; Christian Hubschneider (FZI Research Center for Information Technology); Michael Weber (FZI); Ruby Moritz (Volkswagen); Fabian Hüger (Volkswagen Group Research); Peter Schlicht (Volkswagen Group Research); Marius Zöllner (FZI)

Mapping of the accepted papers to the phases of the DNN development process

• Leveraging combinatorial testing for safety-critical computer vision datasets

• Improved Noise and Attack Robustness for Semantic Segmentation by Using Multi-Task Training with Self-Supervised Depth Estimation (Oral)

• Robust Semantic Segmentation by Redundant Networks With a Layer-Specific Loss Contribution and Majority Vote (Oral)

• Attentional Bottleneck: Towards an Interpretable Deep Driving Network (Oral)

• Mind the Gap - A Benchmark for Dense Depth Prediction Beyond Lidar (Oral)

• Explaining Autonomous Driving by Learning End-to-End Visual Attention (Oral)

• Detection and Retrieval of Out-of-Distribution Objects in Semantic Segmentation (Oral)

• Generating Socially Acceptable Perturbations for Efficient Evaluation of Autonomous Vehicles

• Self-Supervised Domain Mismatch Estimation for Autonomous Perception

• Using Mixture of Expert Models to Gain Insights into Semantic Segmentation

• Unsupervised Temporal Consistency Metric for Video Segmentation in Highly-Automated Driving

• Evaluating Scalable Bayesian Deep Learning Methods for Robust Computer Vision

• Multivariate Confidence Calibration for Object Detection