Depth Forecasting on Frozen Forecasting Unified Evaluation
0.146Mean Abs. Rel. ErrorDINOv2
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| DINOv2Diffusion Model Samples=10, Context Frames=42025.07 | 0.146 | 0.038 | 0.088 | 0.089 | |
| VideoPrismDiffusion Model Samples=10, Context Frames=42025.07 | 0.155 | 0.041 | 0.093 | 0.098 | |
| VideoMAEDiffusion Model Samples=10, Context Frames=42025.07 | 0.189 | 0.064 | 0.091 | 0.104 | |
| N-WALT 500MDiffusion Model Samples=10, Context Frames=42025.07 | 0.19 | 0.026 | 0.145 | 0.158 | |
| 4DS-eDiffusion Model Samples=10, Context Frames=42025.07 | 0.1937 | 0.065 | 0.096 | 0.086 | |
| VJEPADiffusion Model Samples=10, Context Frames=42025.07 | 0.2 | 0.069 | 0.098 | 0.132 | |
| VideoMAEv2Diffusion Model Samples=10, Context Frames=42025.07 | 0.206 | 0.071 | 0.1 | 0.107 | |
| 4DS-hDiffusion Model Samples=10, Context Frames=42025.07 | 0.224 | 0.077 | 0.107 | 0.1156 | |
| SigLIPDiffusion Model Samples=10, Context Frames=42025.07 | 0.233 | 0.036 | 0.175 | 0.144 | |
| WALT 500MDiffusion Model Samples=10, Context Frames=42025.07 | 0.249 | 0.072 | 0.138 | 0.158 |