MoCha-Stereo: Motif Channel Attention Network for Stereo Matching

TL;DR

MoCha-Stereo introduces a novel neural network that enhances stereo matching accuracy by capturing geometric structures and refining disparity estimates, achieving top performance on KITTI benchmarks.

Contribution

The paper proposes the Motif Channel Correlation Volume and Reconstruction Error Motif Penalty modules to improve edge detail matching and disparity accuracy in stereo matching.

Findings

Achieved 1st place on KITTI-2015 and KITTI-2012 leaderboards.

Demonstrated superior performance in Multi-View Stereo tasks.

Effectively preserves geometric structure information during feature processing.

Abstract

Learning-based stereo matching techniques have made significant progress. However, existing methods inevitably lose geometrical structure information during the feature channel generation process, resulting in edge detail mismatches. In this paper, the Motif Cha}nnel Attention Stereo Matching Network (MoCha-Stereo) is designed to address this problem. We provide the Motif Channel Correlation Volume (MCCV) to determine more accurate edge matching costs. MCCV is achieved by projecting motif channels, which capture common geometric structures in feature channels, onto feature maps and cost volumes. In addition, edge variations in %potential feature channels of the reconstruction error map also affect details matching, we propose the Reconstruction Error Motif Penalty (REMP) module to further refine the full-resolution disparity estimation. REMP integrates the frequency information of typical channel features from the reconstruction error. MoCha-Stereo ranks 1st on the KITTI-2015 and KITTI-2012 Reflective leaderboards. Our structure also shows excellent performance in Multi-View Stereo. Code is avaliable at https://github.com/ZYangChen/MoCha-Stereo.