Stereo Matching on Middlebury half-resolution v3
9.4Bad Pixel Error (0.5)MatchAttentionXL
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| MatchAttentionXLTraining Data=Million-scale2025.10 | 9.4 | 3.71 | 1.89 | |
| Fast-FSD-416Training Data=Million-scale, Max Disparity=4162025.10 | 11.8 | 4.81 | 2.2 | |
| Fast-FSD-192Training Data=Million-scale, Max Disparity=1922025.10 | 13.77 | 7.35 | 4.79 | |
| Monster++-416Training Data=Million-scale, Max Disparity=4162025.10 | 16.06 | 6.22 | 2.13 | |
| Zero-RAFT-StereoTraining Data=Million-scale2025.10 | 16.74 | 8.32 | 4.45 | |
| MatchAttentionRTTraining Data=Million-scale2025.10 | 17.1 | 7.55 | 3.76 | |
| RT-IGEV-416Training Data=Scene Flow, Max Disparity=4162025.10 | 25.6 | 15.57 | 9.76 | |
| RT-Monster++-416Training Data=Million-scale, Max Disparity=4162025.10 | 26.34 | 13.1 | 6.16 | |
| RT-IGEV-192Training Data=Scene Flow, Max Disparity=1922025.10 | 26.49 | 16.95 | 11.52 | |
| MatchAttentionRTTraining Data=Scene Flow2025.10 | 27.04 | 14.29 | 8.01 | |
| RT-Monster++-192Training Data=Million-scale, Max Disparity=1922025.10 | 27.1 | 14.24 | 7.43 | |
| LiteAnyStereoTraining Data=Million-scale2025.10 | 36 | 19.02 | 10.39 | |
| IINetTraining Data=Scene Flow2025.10 | 41.31 | 25.64 | 16.37 | |
| LiteAnyStereoTraining Data=Scene Flow2025.10 | 45.63 | 27.27 | 15.75 | |
| LightStereo-LTraining Data=Scene Flow2025.10 | 56.33 | 37.27 | 23.89 | |
| BANet-2DTraining Data=Scene Flow2025.10 | 63.15 | 43.78 | 28.46 | |
| BANet-3D-192Training Data=Scene Flow, Max Disparity=1922025.10 | 63.3 | 44.91 | 30.12 | |
| BANet-3D-416Training Data=Scene Flow, Max Disparity=4162025.10 | 64.17 | 46.16 | 31.42 |