Single-image novel-view video generation on View-of-Delft (test)
14.69PSNRGEN3C (Ours, Multi-Modal with LiDAR)
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| GEN3C (Ours, Multi-Modal with LiDAR)Pixel coverage=0.52%, Sensor=LiDAR2026.02 | 14.69 | 0.4971 | 0.423 | 71.91 | 0.0563 | |
| GEN3C (Ours, Multi-Modal with Radar)Pixel coverage=0.02%, Sensor=Radar2026.02 | 14.26 | 0.486 | 0.4441 | 82.41 | 0.079 | |
| GEN3C (Vision-Only Monocular Depth)Pixel coverage=—, Depth Source=MoGe2026.02 | 12.36 | 0.4561 | 0.5804 | 152.62 | 0.1117 |