Consistency-aware Self-Training for Iterative-based Stereo Matching

Benchmark Model Rank Results
stereo-depth-estimation-on-eth3d-low-res-two-view-stereo-test-nonoccCST-Stereo (extra data)–Bad 1.0: 1.02
stereo-depth-estimation-on-eth3d-stereo-zero-shotCST-RAFT-Stereo (zero-shot) (extra data)–BP-1: 2.40
stereo-depth-estimation-on-kitti-2015-stereo-testCST-Stereo (extra data)–D1-all: 1.50
stereo-depth-estimation-on-middlebury-2014-stereo-zero-shot-half-nonoccCST-Selective-IGEV (zero-shot) (extra data)–BP-2: 6.47