Depth estimation on Goliath UpperBody 47 (12 cameras, 16 frames, 4 subjects)
0.028RMSESapiens-2B
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Sapiens-2BEvaluation Protocol=Our Eval2026.03 | 0.028 | 0.6 | |
| THFMDecoder Training=false, 4 Dense Task=false, 2 Sparse Task=false, Token Mask=false2026.03 | 0.028 | 0.6 | |
| Sapiens-0.3BEvaluation Protocol=Our Eval2026.03 | 0.038 | 0.9 | |
| DAVID-LargeEvaluation Protocol=Our Eval2026.03 | 0.038 | 0.9 | |
| THFMDecoder Training=true, 4 Dense Task=true, 2 Sparse Task=false, Token Mask=false2026.03 | 0.038 | 0.9 | |
| THFMDecoder Training=true, 4 Dense Task=true, 2 Sparse Task=true, Token Mask=true2026.03 | 0.038 | 0.9 | |
| DAVID-BaseEvaluation Protocol=Our Eval2026.03 | 0.043 | 1.1 | |
| Depth-ProEvaluation Protocol=Our Eval2026.03 | 0.06 | 1.5 | |
| Sapiens-2BEvaluation Protocol=Results from DAVID2026.03 | 0.204 | 0.5 | |
| DAVID-LargeEvaluation Protocol=Results from DAVID2026.03 | 0.283 | 0.8 | |
| DAVID-BaseEvaluation Protocol=Results from DAVID2026.03 | 0.316 | 0.9 | |
| Sapiens-0.3BEvaluation Protocol=Results from DAVID2026.03 | 0.368 | 1 | |
| Depth-ProEvaluation Protocol=Results from DAVID2026.03 | 0.442 | 1 | |
| DepthAnythingV2-LEvaluation Protocol=Results from DAVID2026.03 | 0.492 | 1.4 | |
| MiDaS-DPT_LEvaluation Protocol=Results from DAVID2026.03 | 0.553 | 1.5 |