Hand-Object Segmentation on HOI4D mini
97.79Left Hand AccuracyCaRe-Ego
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| CaRe-EgoArchitecture Type=Transformer-based2026.02 | 97.79 | 93.09 | 49.35 | 68.01 | 77.06 | |
| UperNetArchitecture Type=Convolution-based2026.02 | 97.71 | 86.04 | 25.77 | 36.58 | 61.53 | |
| Mask2FormerArchitecture Type=Transformer-based2026.02 | 97.48 | 89.38 | 45.22 | 74.17 | 76.56 | |
| MaskFormerArchitecture Type=Transformer-based2026.02 | 96.63 | 87.83 | 44.81 | 73.47 | 75.69 | |
| InterFormerArchitecture Type=Transformer-based2026.02 | 96.55 | 91.71 | 59.75 | 74.7 | 80.68 | |
| ANNEXEArchitecture Type=MLLM-based2026.02 | 96.54 | 93.18 | 48.77 | 75.19 | 78.42 | |
| SCTNetArchitecture Type=Transformer-based2026.02 | 95.25 | 71.27 | 22.68 | 29.98 | 54.9 | |
| Multi-UNetArchitecture Type=Convolution-based2026.02 | 94.37 | 70.25 | 30.41 | 47.65 | 60.67 | |
| SegformerArchitecture Type=Transformer-based2026.02 | 92.13 | 72.42 | 5.52 | 13.41 | 45.87 | |
| SeqArchitecture Type=Transformer-based2026.02 | 40.9 | 38.05 | 28.99 | 61.67 | 42.4 |