Robust Motion Segmentation From Pairwise Matches

@article{Arrigoni2019RobustMS,
  title={Robust Motion Segmentation From Pairwise Matches},
  author={Federica Arrigoni and Tom{\'a}s Pajdla},
  journal={2019 IEEE/CVF International Conference on Computer Vision (ICCV)},
  year={2019},
  pages={671-681}
}
  • F. Arrigoni, T. Pajdla
  • Published 22 May 2019
  • Computer Science
  • 2019 IEEE/CVF International Conference on Computer Vision (ICCV)
In this paper we consider the problem of motion segmentation, where only pairwise correspondences are assumed as input without prior knowledge about tracks. The problem is formulated as a two-step process. First, motion segmentation is performed on image pairs independently. Secondly, we combine independent pairwise segmentation results in a robust way into the final globally consistent segmentation. Our approach is inspired by the success of averaging methods. We demonstrate in simulated as… 
Motion Segmentation with Pairwise Matches and Unknown Number of Motions
TLDR
This paper addresses motion segmentation, that is the problem of clustering points in multiple images according to a number of moving objects, based on principles from “multi-model fitting” and “permutation synchronization”, and can handle an unknown number of motions.
Multi-frame Motion Segmentation by Combining Two-Frame Results
TLDR
This paper introduces a two-stage approach to motion segmentation on sparse and unstructured datasets involving rigid motions, motivated by multibody structure from motion, and assumes only two-frame correspondences as input without prior knowledge about trajectories.
On the Usage of the Trifocal Tensor in Motion Segmentation
TLDR
This paper addresses motion segmentation in multiple images by combining partial results coming from triplets of images, which are obtained by fitting a number of trifocal tensors to correspondences, exploiting the fact that the trifoc tensor is a stronger model than the fundamental matrix.
Learning To Segment Dominant Object Motion From Watching Videos
TLDR
Inspired by a layered image representation, a technique to group pixel regions according to their affine parametric motion is introduced, which enables the network to learn segmentation of the dominant foreground object using only RGB image pairs as input for both training and inference.
Reconstructing Small 3D Objects in front of a Textured Background
TLDR
In experiments with real artifacts, it is shown that the approach has practical advantages when reconstructing 3D objects from all sides and outperforms the state of the art 3D reconstruction pipeline COLMAP.
Quantum Motion Segmentation
TLDR
This paper introduces the first algorithm for motion segmentation that relies on adiabatic quantum optimization of the objective function and achieves on-par performance with the state of the art on problem instances which can be mapped to modern quantum annealers.

References

SHOWING 1-10 OF 65 REFERENCES
Robust Motion Segmentation with Unknown Correspondences
TLDR
This paper introduces an approach to performing motion segmentation without any prior knowledge of point correspondences in terms of Partial Permutation Matrices (PPMs) and aims to match feature descriptors while simultaneously encouraging point trajectories to satisfy subspace constraints.
Perspective Motion Segmentation via Collaborative Clustering
TLDR
This paper first formulates the 3-D motion segmentation from two perspective views as a subspace clustering problem, utilizing the epipolar constraint of an image pair, and proposes an over-segment and merge approach, where the merging step is based on the property of the ell_1-norm of the mutual sparse representation of two over-Segmented groups.
Robust Realtime Motion-Split-And-Merge for Motion Segmentation
TLDR
Applying RMSAM in a multi-frame motion segmentation context to the Hopkins 155 benchmark, it is shown that compared to the original formulation, the error decreases from 2.05% to only 0.65% at a runtime reduced by 72%.
Geometric motion segmentation and model selection
  • P. Torr
  • Computer Science
    Philosophical Transactions of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences
  • 1998
TLDR
This paper places the problems of model selection, robust estimation and clustering into a common statistical framework; investigating the use of information criteria and robust mixture models as a principled way for motion segmentation of images.
A Benchmark for the Comparison of 3-D Motion Segmentation Algorithms
  • Roberto Tron, R. Vidal
  • Computer Science
    2007 IEEE Conference on Computer Vision and Pattern Recognition
  • 2007
TLDR
This paper compares four 3D motion segmentation algorithms for affine cameras on a benchmark of 155 motion sequences of checkerboard, traffic, and articulated scenes.
Motion Segmentation in the Presence of Outlying, Incomplete, or Corrupted Trajectories
TLDR
A robust subspace separation scheme is developed that deals with practical issues in a unified mathematical framework and gives surprisingly good performance in the presence of the three types of pathological trajectories mentioned above.
Practical Motion Segmentation for Urban Street View Scenes
TLDR
A new approach for image-based motion segmentation in the case of vehicles navigating inside an urban environment by exploiting two application-specific factors and constraining the geometry and exploiting known semantic classes in the scene achieves much higher accuracy than previous approaches.
Projective Factorization of Multiple Rigid-Body Motions
TLDR
The algorithm for segmenting the correspondences according to the multiple motions of point correspondences in multiple perspective views of a scene containing multiple rigid-body motions compares favorably against existing affine motion segmentation methods in terms of segmentation error and execution time.
Motion Segmentation by Exploiting Complementary Geometric Models
TLDR
A multi-view spectral clustering framework that synergistically combines multiple models together that can be substantially improved on existing motion segmentation datasets, achieving state-of-the-art performance on all of them.
Joint estimation of segmentation and structure from motion
...
...