Dichotomous Image Segmentation on DIS TE4 500 (test)
91.2FmaxMVANet
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| MVANetBackbone=Swin-B, Params=93M2025.03 | 91.2 | 85.7 | 94.4 | 90.3 | 0.041 | |
| PDFNetBackbone=Swin-B, Params=94M2025.03 | 91.2 | 86.9 | 94.1 | 91 | 0.037 | |
| DiffDISBackbone=SD-Turbo, Params=865M+2025.03 | 90.9 | 89.3 | 95.5 | 89.6 | 0.025 | |
| BiRefNetBackbone=Swin-L, Params=215M2025.03 | 90.6 | 86.4 | 93.9 | 90 | 0.039 | |
| InSPyReNetBackbone=Swin-B, Params=88M2025.03 | 90.5 | 84.8 | 93.6 | 90.5 | 0.042 | |
| MVANet*Backbone=Swin-B, Params=93M, Input=RGB-depth concatenation2025.03 | 90.5 | 85.6 | 93.2 | 90 | 0.04 | |
| CPNetBackbone=Swin-B & Swin-B, Params=216M2025.03 | 89.5 | 85.6 | 93.5 | 89.6 | 0.039 | |
| GenPerceptBackbone=SD U-Net, Params=865M+2025.03 | 87.4 | 85.8 | 94.7 | 87.4 | 0.041 | |
| MAGNetBackbone=SMT-T & MbNet-V2, Params=16M2025.03 | 87 | 82 | 92.3 | 87 | 0.049 | |
| DACODBackbone=ResNet-50 & Swin-B, Params=189M2025.03 | 86.9 | 81.2 | 91.5 | 87.3 | 0.053 | |
| FP-DIS2025.03 | 84.6 | 78.8 | 90.6 | 85.2 | 0.061 | |
| UDUNBackbone=ResNet-50, Params=25M2025.03 | 84.6 | 79.2 | 90.1 | 84.9 | 0.059 | |
| IS-NetBackbone=U2-Net, Params=44M2025.03 | 82.7 | 75.3 | 87 | 83 | 0.072 | |
| RISNetBackbone=PVT-V2-B2, Params=25M2025.03 | 82.4 | 77.6 | 89 | 85.9 | 0.044 |