Motion Segmentation by Exploiting Complementary Geometric Models

TL;DR

This paper introduces a multi-view spectral clustering framework that combines fundamental matrix and homography models to improve motion segmentation, demonstrating state-of-the-art results on multiple datasets including a new challenging real-world dataset.

Contribution

The paper proposes a novel multi-model spectral clustering approach that effectively integrates fundamental matrix and homography models for enhanced motion segmentation.

Findings

Achieved state-of-the-art performance on existing datasets.

Developed a new challenging dataset based on KITTI with real-world effects.

Demonstrated significant performance improvements over traditional methods.

Abstract

Many real-world sequences cannot be conveniently categorized as general or degenerate; in such cases, imposing a false dichotomy in using the fundamental matrix or homography model for motion segmentation would lead to difficulty. Even when we are confronted with a general scene-motion, the fundamental matrix approach as a model for motion segmentation still suffers from several defects, which we discuss in this paper. The full potential of the fundamental matrix approach could only be realized if we judiciously harness information from the simpler homography model. From these considerations, we propose a multi-view spectral clustering framework that synergistically combines multiple models together. We show that the performance can be substantially improved in this way. We perform extensive testing on existing motion segmentation datasets, achieving state-of-the-art performance on all of them; we also put forth a more realistic and challenging dataset adapted from the KITTI benchmark, containing real-world effects such as strong perspectives and strong forward translations not seen in the traditional datasets.