Metric Depth Estimation on NuScenes
0.051AbsRelTRIG
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| TRIG2026.07 | 0.051 | — | — | 97.4 | |
| DVGT2026.07 | 0.069 | — | — | 95.3 | |
| Ours_CSTM_imageZero-shot Protocol=True2023.07 | 0.147 | 5.889 | — | — | |
| Ours_CSTM_labelZero-shot Protocol=True2023.07 | 0.154 | 7.097 | — | — | |
| OmniVGGT2026.07 | 0.175 | — | — | 44 | |
| Driv3R2026.07 | 0.189 | — | — | 72.1 | |
| VGGT2026.07 | 0.243 | — | — | 72.9 | |
| CUT3R2026.07 | 0.332 | — | — | 54.7 | |
| NewCRFsPre-training Dataset=KITTI, Zero-shot Protocol=True2023.07 | 0.4 | 12.139 | — | — | |
| StreamVGGT2026.07 | 0.412 | — | — | 54 | |
| AdabinsPre-training Dataset=KITTI, Zero-shot Protocol=True2023.07 | 0.445 | 10.658 | — | — | |
| MapAnything2026.07 | 0.568 | — | — | 26.9 | |
| DepthLM-3bModel Type=VLM Trained on Metric Depth Estimation2026.05 | — | — | 87 | — | |
| DepthLM-7bModel Type=VLM Trained on Metric Depth Estimation2026.05 | — | — | 86.5 | — | |
| Gemini-2.5-ProModel Type=VLM2026.05 | — | — | 36.5 | — | |
| GPT-5Model Type=VLM2026.05 | — | — | 38.2 | — | |
| Qwen2.5-VL-72bModel Type=VLM2026.05 | — | — | 18.6 | — | |
| Qwen3-VL-32bModel Type=VLM2026.05 | — | — | 2.9 | — | |
| Qwen3-VL-4bModel Type=VLM2026.05 | — | — | 7 | — | |
| Seed1.5-VLModel Type=VLM Trained on Metric Depth Estimation2026.05 | — | — | 2.8 | — | |
| SpaceLLaVA-13BModel Type=Spatial VLM2026.05 | — | — | 8.3 | — | |
| SpatialRGPT-8BModel Type=Spatial VLM2026.05 | — | — | 10 | — | |
| VLM3-4bModel Type=VLM Trained on Metric Depth Estimation2026.05 | — | — | 97 | — |