18 citations · 49 across the 12 of their papers we have counts for
6 papers · 1 filter
High Resolution Multi-Scale RAFT (Robust Vision Challenge 2022)
Azin Jahedi, Maximilian Luz, Lukas Mehl +2
In this report, we present our optical flow approach, MS-RAFT+, that won the Robust Vision Challenge 2022. It is based on the MS-RAFT method, which successfully integrates several…
Attacking Motion Estimation with Adversarial Snow
Jenny Schmalfuss, Lukas Mehl, Andrés Bruhn
Current adversarial attacks for motion estimation (optical flow) optimize small per-pixel perturbations, which are unlikely to appear in the real world. In contrast, we exploit a r…
Multi-Scale RAFT: Combining Hierarchical Concepts for Learning-based Optical FLow Estimation
Azin Jahedi, Lukas Mehl, Marc Rivinius +1
Many classical and learning-based optical flow methods rely on hierarchical concepts to improve both accuracy and robustness. However, one of the currently most successful approach…
M-FUSE: Multi-frame Fusion for Scene Flow Estimation
Lukas Mehl, Azin Jahedi, Jenny Schmalfuss +1
Recently, neural network for scene flow estimation show impressive results on automotive data such as the KITTI benchmark. However, despite of using sophisticated rigidity assumpti…
Blind Image Inpainting with Sparse Directional Filter Dictionaries for Lightweight CNNs
Jenny Schmalfuss, Erik Scheurer, Heng Zhao +3
Blind inpainting algorithms based on deep learning architectures have shown a remarkable performance in recent years, typically outperforming model-based methods both in terms of i…
A Perturbation-Constrained Adversarial Attack for Evaluating the Robustness of Optical Flow
Jenny Schmalfuss, Philipp Scholze, Andrés Bruhn
Recent optical flow methods are almost exclusively judged in terms of accuracy, while their robustness is often neglected. Although adversarial attacks offer a useful tool to perfo…