View-Conditioned Spatial Reasoning on ViewBench Text
100AccuracyTarget View (Oracle)
Evaluation Results
| Method | Links | |
|---|---|---|
| Target View (Oracle)View Overlap (%)=5–15, Depth=GT2026.04 | 100 | |
| Target View (Oracle)View Overlap (%)=15–25, Depth=GT2026.04 | 100 | |
| Target View (Oracle)View Overlap (%)=25–35, Depth=GT2026.04 | 100 | |
| Backward-Nearest Token WarpingView Overlap (%)=15–25, Depth=Pred.2026.04 | 81.73 | |
| Backward-Nearest Token WarpingView Overlap (%)=15–25, Depth=GT2026.04 | 80.71 | |
| Backward-Adaptive Token WarpingView Overlap (%)=15–25, Depth=Pred.2026.04 | 80.71 | |
| Backward-Adaptive Token WarpingView Overlap (%)=15–25, Depth=GT2026.04 | 79.7 | |
| Backward-Adaptive Token WarpingView Overlap (%)=25–35, Depth=GT2026.04 | 78.86 | |
| Backward-Adaptive Token WarpingView Overlap (%)=5–15, Depth=GT2026.04 | 77.89 | |
| Backward-Nearest Token WarpingView Overlap (%)=25–35, Depth=Pred.2026.04 | 76 | |
| Pixel-Wise Warping (Backward)View Overlap (%)=15–25, Depth=GT2026.04 | 75.63 | |
| Backward-Nearest Token WarpingView Overlap (%)=5–15, Depth=Pred.2026.04 | 75.38 | |
| Backward-Nearest Token WarpingView Overlap (%)=5–15, Depth=GT2026.04 | 74.87 | |
| Backward-Nearest Token WarpingView Overlap (%)=25–35, Depth=GT2026.04 | 74.86 | |
| Pixel-Wise Warping (Backward)View Overlap (%)=15–25, Depth=Pred.2026.04 | 74.62 | |
| Backward-Adaptive Token WarpingView Overlap (%)=25–35, Depth=Pred.2026.04 | 74.29 | |
| Pixel-Wise Warping (Forward)View Overlap (%)=15–25, Depth=GT2026.04 | 73.6 | |
| Backward-Adaptive Token WarpingView Overlap (%)=5–15, Depth=Pred.2026.04 | 73.37 | |
| Pixel-Wise Warping (Forward)View Overlap (%)=15–25, Depth=Pred.2026.04 | 73.1 | |
| Pixel-Wise Warping (Backward)View Overlap (%)=5–15, Depth=GT2026.04 | 71.86 | |
| GenWarpView Overlap (%)=15–25, Depth=GT2026.04 | 71.07 | |
| Pixel-Wise Warping (Forward)View Overlap (%)=5–15, Depth=GT2026.04 | 70.85 | |
| VLM-3RView Overlap (%)=15–25, Depth=GT2026.04 | 70.56 | |
| GenWarpView Overlap (%)=5–15, Depth=GT2026.04 | 69.35 | |
| Pixel-Wise Warping (Forward)View Overlap (%)=5–15, Depth=Pred.2026.04 | 69.35 | |
| Pixel-Wise Warping (Backward)View Overlap (%)=25–35, Depth=GT2026.04 | 68.57 | |
| Pixel-Wise Warping (Backward)View Overlap (%)=25–35, Depth=Pred.2026.04 | 68.57 | |
| Pixel-Wise Warping (Backward)View Overlap (%)=5–15, Depth=Pred.2026.04 | 67.84 | |
| Pixel-Wise Warping (Forward)View Overlap (%)=25–35, Depth=Pred.2026.04 | 67.43 | |
| Forward Token WarpingView Overlap (%)=5–15, Depth=Pred.2026.04 | 66.83 | |
| GenWarpView Overlap (%)=25–35, Depth=GT2026.04 | 66.29 | |
| Forward Token WarpingView Overlap (%)=15–25, Depth=Pred.2026.04 | 65.48 | |
| Forward Token WarpingView Overlap (%)=15–25, Depth=GT2026.04 | 64.47 | |
| VLM-3RView Overlap (%)=5–15, Depth=GT2026.04 | 63.82 | |
| Pixel-Wise Warping (Forward)View Overlap (%)=25–35, Depth=GT2026.04 | 62.86 | |
| VLM-3RView Overlap (%)=25–35, Depth=GT2026.04 | 60.57 | |
| Forward Token WarpingView Overlap (%)=25–35, Depth=Pred.2026.04 | 60.57 | |
| Forward Token WarpingView Overlap (%)=5–15, Depth=GT2026.04 | 60.3 | |
| Qwen2.5-VLView Overlap (%)=15–25, Depth=GT2026.04 | 59.39 | |
| Forward Token WarpingView Overlap (%)=25–35, Depth=GT2026.04 | 54.86 | |
| SpatialReasonerView Overlap (%)=25–35, Depth=GT2026.04 | 53.71 | |
| SpatialReasonerView Overlap (%)=15–25, Depth=GT2026.04 | 53.3 | |
| ViLaSRView Overlap (%)=15–25, Depth=GT2026.04 | 52.28 | |
| Qwen2.5-VLView Overlap (%)=25–35, Depth=GT2026.04 | 52 | |
| ViLaSRView Overlap (%)=25–35, Depth=GT2026.04 | 48 | |
| SpatialReasonerView Overlap (%)=5–15, Depth=GT2026.04 | 46.73 | |
| Qwen2.5-VLView Overlap (%)=5–15, Depth=GT2026.04 | 46.23 | |
| ViLaSRView Overlap (%)=5–15, Depth=GT2026.04 | 44.22 |