Medical Image Segmentation on MosMedData+
68.53mIoUCERS
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| CERSText=Yes, Backbone=ConvNeXt, Labeled ratio=100%2026.06 | 68.53 | 79.21 | — | — | — | |
| TGC-NetText-enhanced=true2025.12 | 68 | 80.94 | — | — | — | |
| BiCLIPText-Guided=Yes, Backbone=CNN2026.02 | 67.79 | 80.8 | — | — | — | |
| EF-UNetVenue=arXiv’25, Text-Guided=Yes, Backbone=CNN2026.02 | 67.37 | 80.5 | — | — | — | |
| CERSText=Yes, Backbone=ConvNeXt, Labeled ratio=50%2026.06 | 66.87 | 78.26 | — | — | — | |
| LanGuideSegBackbone=CNN, Params ↓ (M)=64.35, FLOPs ↓ (G)=36.27, Multi-modal=true2026.04 | 66.05 | 78.11 | — | — | — | |
| U-NetText=No, Backbone=ConvNeXt, Labeled ratio=100%2026.06 | 66.03 | 76.99 | — | — | — | |
| FMISegBackbone=CNN, Text=Yes2025.09 | 65.71 | 79.3 | — | — | — | |
| TAMISegBackbone=CNN, Params ↓ (M)=48.75, FLOPs ↓ (G)=29.83, Multi-modal=true2026.04 | 65.71 | 79.3 | — | — | — | |
| TGCAMBackbone=CNN, Params ↓ (M)=61.76, FLOPs ↓ (G)=35.99, Multi-modal=true2026.04 | 65.32 | 77.73 | — | — | — | |
| ConDSegBackbone=CNN, Params ↓ (M)=43.81, FLOPs ↓ (G)=28.62, Multi-modal=false2026.04 | 65.18 | 75.89 | — | — | — | |
| RecLMISBackbone=CNN, Text=Yes2025.09 | 65.07 | 77.48 | — | — | — | |
| RecLMISVenue=TMI’24, Text-enhanced=true2025.12 | 65.07 | 77.48 | — | — | — | |
| RecLMISVenue=IEEE TMI’24, Text-Guided=Yes, Backbone=CNN2026.02 | 65.07 | 77.48 | — | — | — | |
| TVE-NetBackbone=CNN, Params ↓ (M)=56.26, FLOPs ↓ (G)=33.46, Multi-modal=true2026.04 | 65.07 | 77.48 | — | — | — | |
| MAdapterBackbone=CNN, Text=Yes2025.09 | 64.78 | 78.62 | — | — | — | |
| MAdapterBackbone=CNN, Params ↓ (M)=67.55, FLOPs ↓ (G)=37.65, Multi-modal=true2026.04 | 64.78 | 78.62 | — | — | — | |
| UCMTText=No, Backbone=ConvNeXt, Labeled ratio=50%2026.06 | 64.65 | 75.95 | — | — | — | |
| MedSegDiff-V2Backbone=Diffusion, Params ↓ (M)=46.12, FLOPs ↓ (G)=149.01, Multi-modal=false2026.04 | 64.55 | 78.59 | — | — | — | |
| SMF-netVenue=Frontiers’25, Text-Guided=Yes, Backbone=Hybrid2026.02 | 64.17 | 78.7 | — | — | — | |
| VT-MFLVVenue=Imaging’25, Text-Guided=Yes, Backbone=CNN2026.02 | 63.98 | 75.61 | — | — | — | |
| TGCAMBackbone=CNN, Text=Yes2025.09 | 63.69 | 77.82 | — | — | — | |
| ProLearnBackbone=CNN, Params ↓ (M)=69.88, FLOPs ↓ (G)=38.12, Multi-modal=true2026.04 | 63.69 | 77.82 | — | — | — | |
| Ariadne’s ThreadVenue=MICCAI’23, Text-enhanced=true2025.12 | 63.6 | 77.75 | — | — | — | |
| CERSText=Yes, Backbone=ConvNeXt, Labeled ratio=25%2026.06 | 63.39 | 74.73 | — | — | — | |
| LeFeDText=No, Backbone=V-Net [12], Labeled ratio=50%2026.06 | 63.38 | 74.77 | — | — | — | |
| Qwen2.5VL-3BModel size=3B2026.05 | 63.33 | 76.23 | — | — | — | |
| LViTText=Yes, Backbone=LViT, Labeled ratio=50%2026.06 | 63.27 | 75.47 | — | — | — | |
| MedLangViTVenue=Electronics’25, Text-Guided=Yes, Backbone=Hybrid2026.02 | 63.17 | 75.95 | — | — | — | |
| DuCiSCText=No, Backbone=ConvNeXt, Labeled ratio=50%2026.06 | 63.08 | 74.24 | — | — | — | |
| CausalCLIPSegBackbone=Hybrid, Params ↓ (M)=89.86, FLOPs ↓ (G)=45.95, Multi-modal=true2026.04 | 63.05 | 76.12 | — | — | — | |
| Cap2SegVenue=Frontiers’24, Text-Guided=Yes, Backbone=Hybrid2026.02 | 63.02 | 75.87 | — | — | — | |
| TextDiffBackbone=Diffusion, Params ↓ (M)=9.68, FLOPs ↓ (G)=36.85, Multi-modal=true2026.04 | 62.77 | 76.52 | — | — | — | |
| TIRNetParams (M)=41.10, Flops (G)=24.1, Modality=multi-modal2026.06 | 62.77 | — | 75.41 | 81.38 | 68.61 | |
| LGABackbone=Transformer, Text=Yes2025.09 | 62.52 | 75.63 | — | — | — | |
| LGAVenue=MICCAI’24, Text-enhanced=true2025.12 | 62.52 | 75.63 | — | — | — | |
| LGAVenue=MICCAI’24, Text-Guided=Yes, Backbone=ViT2026.02 | 62.52 | 75.63 | — | — | — | |
| LGABackbone=Transformer, Params ↓ (M)=97.58, FLOPs ↓ (G)=48.23, Multi-modal=true2026.04 | 62.52 | 75.63 | — | — | — | |
| Janus1BModel size=1B2026.05 | 62.39 | 75.46 | — | — | — | |
| MTText=No, Backbone=ConvNeXt, Labeled ratio=50%2026.06 | 61.97 | 73.44 | — | — | — | |
| RefSegformerVenue=TIP’24, Text-enhanced=true2025.12 | 61.7 | 74.98 | — | — | — | |
| LViT-TBackbone=Hybrid, Text=Yes2025.09 | 61.33 | 74.57 | — | — | — | |
| LVITVenue=TMI’23, Text-enhanced=true2025.12 | 61.33 | 74.57 | — | — | — | |
| LViTVenue=IEEE TMI’23, Text-Guided=Yes, Backbone=Hybrid2026.02 | 61.33 | 74.57 | — | — | — | |
| LViTBackbone=Hybrid, Params ↓ (M)=93.16, FLOPs ↓ (G)=46.75, Multi-modal=true2026.04 | 61.33 | 74.57 | — | — | — | |
| LViT-T2026.05 | 61.33 | 74.57 | — | — | — | |
| LViTParams (M)=29.70, Flops (G)=54.1, Modality=multi-modal2026.06 | 61.15 | — | 74.52 | 76.64 | 62.12 | |
| LGAParams (M)=8.24, Flops (G)=381.1, Modality=multi-modal2026.06 | 61.1 | — | 74.53 | 80.55 | 67.43 | |
| nnUNetParams (M)=19.10, Flops (G)=412.7, Modality=uni-modal2026.06 | 61.02 | — | 73.18 | 80.46 | 67.31 | |
| BCPText=No, Backbone=ConvNeXt, Labeled ratio=50%2026.06 | 60.79 | 72.57 | — | — | — | |
| LAVTParams (M)=118.60, Flops (G)=83.8, Modality=multi-modal2026.06 | 60.74 | — | 73.66 | 80.66 | 67.58 | |
| UCMTText=No, Backbone=ConvNeXt, Labeled ratio=25%2026.06 | 60.47 | 72 | — | — | — | |
| LAVTVenue=CVPR’22, Text-enhanced=true2025.12 | 60.41 | 73.29 | — | — | — | |
| LAVTVenue=CVPR’22, Text-Guided=Yes, Backbone=Hybrid2026.02 | 60.41 | 73.29 | — | — | — | |
| LAVT2026.05 | 60.41 | 73.29 | — | — | — | |
| nnUnetBackbone=CNN, Text=No2025.09 | 60.36 | 72.59 | — | — | — | |
| nnUNetVenue=Nature’21, Text-enhanced=false2025.12 | 60.36 | 72.59 | — | — | — | |
| nnU-NetVenue=Nature’21, Text-Guided=No, Backbone=CNN2026.02 | 60.36 | 72.59 | — | — | — | |
| nnU-NetBackbone=CNN, Params ↓ (M)=41.25, FLOPs ↓ (G)=32.71, Multi-modal=false2026.04 | 60.36 | 72.59 | — | — | — | |
| nnUNet2026.05 | 60.36 | 72.59 | — | — | — | |
| DuCiSCText=No, Backbone=ConvNeXt, Labeled ratio=25%2026.06 | 60.3 | 71.67 | — | — | — | |
| CMIRNetParams (M)=239.80, Flops (G)=134.5, Modality=multi-modal2026.06 | 60.25 | — | 73.88 | 79.64 | 66.17 | |
| GLoRIAVenue=ICCV’21, Text-enhanced=true2025.12 | 60.18 | 72.42 | — | — | — | |
| GLoRIA2026.05 | 60.18 | 72.42 | — | — | — | |
| ViLTVenue=ICML’21, Text-enhanced=true2025.12 | 60.15 | 72.36 | — | — | — | |
| ViLT2026.05 | 60.15 | 72.36 | — | — | — | |
| ARsegVenue=MICCAI’25, Text-enhanced=true2025.12 | 59.82 | 73.24 | — | — | — | |
| ARSegVenue=MICCAI’25, Text-Guided=Yes, Backbone=Hybrid2026.02 | 59.82 | 73.24 | — | — | — | |
| ConVIRT2026.05 | 59.73 | 72.06 | — | — | — | |
| BCPText=No, Backbone=ConvNeXt, Labeled ratio=25%2026.06 | 59.37 | 71.31 | — | — | — | |
| TGANetVenue=MICCAI’22, Text-enhanced=true2025.12 | 59.28 | 71.81 | — | — | — | |
| TGANetVenue=MICCAI’22, Text-Guided=Yes, Backbone=CNN2026.02 | 59.28 | 71.81 | — | — | — | |
| TGANet2026.05 | 59.28 | 71.81 | — | — | — | |
| TeViAParams (M)=146.80, Flops (G)=11.2, Modality=multi-modal2026.06 | 59.06 | — | 72.32 | 79.47 | 65.94 | |
| LeFeDText=No, Backbone=V-Net [12], Labeled ratio=25%2026.06 | 58.95 | 70.45 | — | — | — | |
| ATParams (M)=44.00, Flops (G)=22.4, Modality=multi-modal2026.06 | 58.63 | — | 72.41 | 78.53 | 64.65 | |
| TransUNetBackbone=Hybrid, Text=No2025.09 | 58.44 | 71.24 | — | — | — | |
| TransUNetBackbone=Hybrid, Params ↓ (M)=104.93, FLOPs ↓ (G)=37.22, Multi-modal=false2026.04 | 58.44 | 71.24 | — | — | — | |
| TransUNet2026.05 | 58.44 | 71.24 | — | — | — | |
| UNet++Backbone=CNN, Text=No2025.09 | 58.39 | 71.75 | — | — | — | |
| U-Net++Venue=MICCAI’18, Text-enhanced=false2025.12 | 58.39 | 71.75 | — | — | — | |
| UNet++Venue=MICCAI’18, Text-Guided=No, Backbone=CNN2026.02 | 58.39 | 71.75 | — | — | — | |
| U-Net++Backbone=CNN, Params ↓ (M)=36.01, FLOPs ↓ (G)=25.22, Multi-modal=false2026.04 | 58.39 | 71.75 | — | — | — | |
| MTText=No, Backbone=ConvNeXt, Labeled ratio=25%2026.06 | 58.34 | 70.08 | — | — | — | |
| LViTText=Yes, Backbone=LViT, Labeled ratio=25%2026.06 | 58.13 | 71.16 | — | — | — | |
| TGANetParams (M)=19.80, Flops (G)=41.9, Modality=multi-modal2026.06 | 57.9 | — | 70.73 | 80.06 | 66.75 | |
| U-Net++Params (M)=74.50, Flops (G)=94.6, Modality=uni-modal2026.06 | 57.39 | — | 69.74 | 80.19 | 66.93 | |
| GLoRIAParams (M)=45.60, Flops (G)=60.8, Modality=multi-modal2026.06 | 56.75 | — | 70.11 | 79.31 | 65.71 | |
| UCTransNetBackbone=Hybrid, Text=No2025.09 | 52.69 | 65.9 | — | — | — | |
| UCTransNetVenue=AAAI’22, Text-enhanced=false2025.12 | 52.69 | 65.9 | — | — | — | |
| UCTransNetVenue=AAAI’22, Text-Guided=No, Backbone=Hybrid2026.02 | 52.69 | 65.9 | — | — | — | |
| MAXFormerBackbone=Hybrid, Params ↓ (M)=68.96, FLOPs ↓ (G)=44.87, Multi-modal=false2026.04 | 52.69 | 65.9 | — | — | — | |
| Swin-UNetParams (M)=82.30, Flops (G)=67.3, Modality=uni-modal2026.06 | 52.35 | — | 65.94 | 77.85 | 63.73 | |
| U-NetBackbone=CNN, Text=No2025.09 | 50.73 | 64.6 | — | — | — | |
| U-NetVenue=MICCAI’15, Text-enhanced=false2025.12 | 50.73 | 64.6 | — | — | — | |
| U-NetVenue=MICCAI’15, Text-Guided=No, Backbone=CNN2026.02 | 50.73 | 64.6 | — | — | — | |
| U-NetBackbone=CNN, Params ↓ (M)=31.04, FLOPs ↓ (G)=54.75, Multi-modal=false2026.04 | 50.73 | 64.6 | — | — | — | |
| Swin-UnetBackbone=Hybrid, Text=No2025.09 | 50.19 | 63.29 | — | — | — | |
| Swin-UNetVenue=ECCV’22, Text-enhanced=false2025.12 | 50.19 | 63.29 | — | — | — | |
| Swin-UnetVenue=ECCV’22, Text-Guided=No, Backbone=Hybrid2026.02 | 50.19 | 63.29 | — | — | — |