Polyp Segmentation on Kvasir-SEG (test)
87.1mIoUADSNet
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ADSNet2024.05 | 87.1 | 92 | — | — | — | — | — | 91.6 | 0.02 | — | — | |
| Polyp-PVT2024.05 | 86.4 | 91.7 | — | — | — | — | — | 91.1 | 0.023 | — | — | |
| Phi-SegNet2026.01 | 84.96 | 91.17 | 92.64 | 91.85 | — | 97.61 | — | — | — | 92.24 | 1.7547 | |
| SANet2024.05 | 84.7 | 90.4 | — | — | — | — | — | 89.2 | 0.028 | — | — | |
| AG-CUResNeSt-101Training scenario=Scenario 42021.05 | 84.5 | 90.2 | — | — | — | — | — | — | — | — | — | |
| PraNetTraining scenario=Scenario 42021.05 | 84 | 89.8 | — | — | — | — | — | — | — | — | — | |
| PraNet2024.05 | 84 | 89.8 | — | — | — | — | — | 88.5 | 0.03 | — | — | |
| DCRNet2024.05 | 82.5 | 88.6 | — | — | — | — | — | 86.8 | 0.035 | — | — | |
| CPFNet2026.01 | 81.62 | 87.94 | 90.53 | 90.13 | — | 96.53 | — | — | — | 90.33 | 2.9053 | |
| Twin-SegNet2026.01 | 81.09 | 88.26 | 91.85 | 87.31 | — | 96.48 | — | — | — | 89.34 | 2.2324 | |
| HarDNet-MSEGTraining scenario=Scenario 4, Retrained with original reported configurations=true2021.05 | 80.7 | 87.7 | — | — | — | — | — | — | — | — | — | |
| CE-Net2026.01 | 80.32 | 87.01 | 88.66 | 91.23 | — | 96.37 | — | — | — | 89.82 | 2.5477 | |
| ResUNet++Training scenario=Scenario 42021.05 | 79.3 | 81.3 | — | — | — | — | — | — | — | — | — | |
| U-Net-ResNet50Backbone=ResNet-50, Pre-training=ImageNet2020.12 | 78.71 | 89.26 | 84.33 | 92.07 | — | 96.39 | 85.85 | — | — | — | — | |
| DDANet2020.12 | 78 | 85.76 | 88.8 | 86.43 | 69.59 | — | — | — | — | — | — | |
| MEW-UNet2026.01 | 77.53 | 85.4 | 87.08 | 89.68 | — | 95.16 | — | — | — | 88.35 | 3.1729 | |
| TransAttUNet2026.01 | 77.46 | 85.57 | 89.2 | 87.61 | — | 96.58 | — | — | — | 88.4 | 2.6258 | |
| AAU-Net2026.01 | 75.84 | 83.51 | 86.67 | 86.22 | — | 94.87 | — | — | — | 86.44 | 3.4935 | |
| PraNet2026.01 | 75.5 | 83.77 | 89.32 | 84.9 | — | 94.71 | — | — | — | 87.09 | 3.8235 | |
| UNet++2026.01 | 74.65 | 83.44 | 89.36 | 83.68 | — | 94.9 | — | — | — | 86.42 | 3.1446 | |
| UNetTraining scenario=Scenario 42021.05 | 74.6 | 81.8 | — | — | — | — | — | — | — | — | — | |
| UNet++Training scenario=Scenario 42021.05 | 74.3 | 82.1 | — | — | — | — | — | — | — | — | — | |
| Y-Net2026.01 | 72.54 | 81.7 | 84.11 | 85.95 | — | 94.24 | — | — | — | 85.01 | 3.9857 | |
| ResUNet++2026.01 | 71.4 | 80.14 | 85.38 | 83.98 | — | 93.99 | — | — | — | 84.67 | 3.9315 | |
| DoubleUNetTraining scenario=Scenario 4, Retrained with original reported configurations=true2021.05 | 70 | 78.1 | — | — | — | — | — | — | — | — | — | |
| DDANetTraining scenario=Scenario 4, Retrained with original reported configurations=true2021.05 | 65.8 | 75.8 | — | — | — | — | — | — | — | — | — | |
| UNet2026.01 | 64.33 | 75.66 | 77.54 | 82.11 | — | 92.66 | — | — | — | 79.76 | 5.3399 | |
| ColonSegNetTraining scenario=Scenario 4, Retrained with original reported configurations=true2021.05 | 64.3 | 75.3 | — | — | — | — | — | — | — | — | — | |
| SFATraining scenario=Scenario 42021.05 | 61.1 | 72.3 | — | — | — | — | — | — | — | — | — | |
| M-BFF-Net2025.10 | 0.9296 | 0.9514 | 0.9469 | 0.9738 | — | 0.9878 | — | — | — | — | — | |
| Meta-Polyp2025.10 | 0.921 | 0.959 | 0.9337 | 0.935 | — | 0.9789 | — | — | — | — | — | |
| EffiSegNet-B5Pre-trained on external dataset=true2024.07 | 0.9065 | 0.9488 | 0.9321 | 0.9713 | — | — | — | — | — | 0.9513 | — | |
| EffiSegNet-B6Pre-trained on external dataset=true2024.07 | 0.906 | 0.9477 | 0.9334 | 0.9724 | — | — | — | — | — | 0.9531 | — | |
| EffiSegNet-B4Pre-trained on external dataset=true2024.07 | 0.9056 | 0.9483 | 0.9429 | 0.9679 | — | — | — | — | — | 0.9552 | — | |
| DUCK-Net2024.07 | 0.9051 | — | 0.9379 | 0.9628 | — | — | — | — | — | 0.927 | — | |
| Duck-Net2025.10 | 0.9051 | 0.9502 | 0.9379 | 0.9628 | — | 0.9842 | — | — | — | — | — | |
| SAMed + SAIFCategory=SAM-based, Prompting Strategy=Box prompts, Inference Strategy=SAIF2026.03 | 0.8942 | 0.9296 | — | — | — | — | — | — | — | — | — | |
| MSRF-Net2022.08 | 0.8914 | 0.9217 | 0.9198 | 0.9666 | — | — | — | — | — | — | — | |
| FCBFormer2022.08 | 0.8903 | 0.9385 | 0.9401 | 0.9459 | — | — | — | — | — | — | — | |
| EffiSegNet-B3Pre-trained on external dataset=true2024.07 | 0.8876 | 0.9358 | 0.9321 | 0.9613 | — | — | — | — | — | 0.9465 | — | |
| EffiSegNet-B2Pre-trained on external dataset=true2024.07 | 0.8836 | 0.9329 | 0.938 | 0.955 | — | — | — | — | — | 0.9464 | — | |
| Li-SegPNetPre-trained on external dataset=true2024.07 | 0.88 | — | 0.9254 | 0.9424 | — | — | — | — | — | 0.9058 | — | |
| EffiSegNet-B0Pre-trained on external dataset=true2024.07 | 0.8794 | 0.9304 | 0.9368 | 0.9475 | — | — | — | — | — | 0.9421 | — | |
| SAM2-UNetResolution=352x352, Multi-scale training={1, 1.25}2024.08 | 0.879 | 0.928 | — | — | — | — | — | — | — | — | — | |
| EffiSegNet-B1Pre-trained on external dataset=true2024.07 | 0.8784 | 0.9288 | 0.9461 | 0.9437 | — | — | — | — | — | 0.9448 | — | |
| ColonFormerPre-trained on external dataset=true2024.07 | 0.877 | — | — | — | — | — | — | — | — | 0.9502 | — | |
| EffiSegNet-B4Pre-trained on external dataset=false2024.07 | 0.8668 | 0.9207 | 0.9262 | 0.9311 | — | — | — | — | — | 0.9286 | — | |
| Polyp-PVTCategory=Train-based, Prompting Strategy=Prompt-free2026.03 | 0.8642 | 0.917 | — | — | — | — | — | — | — | — | — | |
| MSNetCategory=Train-based, Prompting Strategy=Prompt-free2026.03 | 0.8622 | 0.9075 | — | — | — | — | — | — | — | — | — | |
| CFA-Net2024.08 | 0.861 | 0.915 | — | — | — | — | — | — | — | — | — | |
| CaraNet2024.08 | 0.859 | 0.913 | — | — | — | — | — | — | — | — | — | |
| TBconvL-Net2025.10 | 0.8554 | 0.922 | 0.9203 | 0.9238 | — | 0.9749 | — | — | — | — | — | |
| EU-NetCategory=Train-based, Prompting Strategy=Prompt-free2026.03 | 0.8542 | 0.9081 | — | — | — | — | — | — | — | — | — | |
| Patch-MoE Mamba2026.05 | 0.8532 | 0.909 | — | — | — | — | — | — | 0.0258 | — | — | |
| U-Net v22026.05 | 0.8531 | 0.9084 | — | — | — | — | — | — | 0.0265 | — | — | |
| VM-UNetV22026.05 | 0.853 | 0.9082 | — | — | — | — | — | — | 0.026 | — | — | |
| SAM Adapter + SAIFCategory=SAM-based, Prompting Strategy=Box prompts, Inference Strategy=SAIF2026.03 | 0.8492 | 0.9133 | — | — | — | — | — | — | — | — | — | |
| PraNetCategory=Train-based, Prompting Strategy=Prompt-free2026.03 | 0.8483 | 0.9015 | — | — | — | — | — | — | — | — | — | |
| SANet2024.08 | 0.847 | 0.904 | — | — | — | — | — | — | — | — | — | |
| PraNet2022.08 | 0.84 | 0.898 | — | — | — | — | — | — | — | — | — | |
| PraNet2024.08 | 0.84 | 0.898 | — | — | — | — | — | — | — | — | — | |
| VM-UNet2026.05 | 0.84 | 0.899 | — | — | — | — | — | — | 0.0294 | — | — | |
| PraNetPre-trained on external dataset=true, Evaluation source=[6]2024.07 | 0.8339 | — | 0.864 | 0.9599 | — | — | — | — | — | 0.9094 | — | |
| UNet++2025.10 | 0.8339 | 0.9094 | 0.864 | 0.9599 | — | 0.9738 | — | — | — | — | — | |
| TGANetBackbone=ResNet502022.05 | 0.833 | 0.8982 | 0.9132 | 0.9123 | — | — | 0.9029 | — | — | — | — | |
| PraNetBackbone=Res2Net2022.05 | 0.8296 | 0.8942 | 0.906 | 0.9126 | — | — | 0.8976 | — | — | — | — | |
| FCBFormerTraining data=CVC-ClinicDB [2]2022.08 | 0.8214 | 0.8848 | 0.8754 | 0.9354 | — | — | — | — | — | — | — | |
| SwinUNetCategory=Train-based, Prompting Strategy=Prompt-free2026.03 | 0.8206 | 0.8892 | — | — | — | — | — | — | — | — | — | |
| DeepLabV3+Backbone=ResNet502022.05 | 0.8172 | 0.8837 | 0.9014 | 0.9028 | — | — | 0.8904 | — | — | — | — | |
| SAMedCategory=SAM-based, Prompting Strategy=Box prompts2026.03 | 0.8161 | 0.8801 | — | — | — | — | — | — | — | — | — | |
| TransUNetCategory=Train-based, Prompting Strategy=Prompt-free2026.03 | 0.8058 | 0.8691 | — | — | — | — | — | — | — | — | — | |
| SAM AdapterCategory=SAM-based, Prompting Strategy=Box prompts2026.03 | 0.8025 | 0.8631 | — | — | — | — | — | — | — | — | — | |
| TransNetRFPS=54.60, Parameters (Millions)=27.27, Flops (GMac)=10.582023.03 | 0.8016 | 0.8706 | 0.8843 | 0.9073 | — | — | 0.8744 | — | — | — | — | |
| ResUNet++2022.08 | 0.7927 | 0.8133 | 0.8774 | 0.7064 | — | — | — | — | — | — | — | |
| ResUNet++2024.07 | 0.7927 | — | 0.8774 | 0.7064 | — | — | — | — | — | 0.8133 | — | |
| ResUNet2024.07 | 0.7778 | — | — | — | — | — | — | — | — | 0.7878 | — | |
| UACANetFPS=25.85, Parameters (Millions)=69.16, Flops (GMac)=31.512023.03 | 0.7692 | 0.8502 | 0.8799 | 0.8706 | — | — | 0.8626 | — | — | — | — | |
| MedSAM + SAIFCategory=SAM-based, Prompting Strategy=Box prompts, Inference Strategy=SAIF2026.03 | 0.7635 | 0.8516 | — | — | — | — | — | — | — | — | — | |
| U-NetEvaluation source=[6]2024.07 | 0.7629 | — | 0.8718 | 0.8593 | — | — | — | — | — | 0.8655 | — | |
| U-Net2025.10 | 0.7629 | 0.8655 | 0.8718 | 0.8593 | — | 0.9563 | — | — | — | — | — | |
| ARU-GD2025.10 | 0.7584 | 0.8626 | 0.8005 | 0.9351 | — | 0.9583 | — | — | — | — | — | |
| U-Net2026.05 | 0.7558 | 0.8319 | — | — | — | — | — | — | 0.0472 | — | — | |
| U-Net2022.05 | 0.7472 | 0.8264 | 0.8504 | 0.8703 | — | — | 0.8353 | — | — | — | — | |
| U-NetFPS=106.88, Parameters (Millions)=31.04, Flops (GMac)=54.752023.03 | 0.7472 | 0.8264 | 0.8504 | 0.8703 | — | — | 0.8353 | — | — | — | — | |
| U-NetCategory=Train-based, Prompting Strategy=Prompt-free2026.03 | 0.7463 | 0.8184 | — | — | — | — | — | — | — | — | — | |
| HarDNet-MSEGBackbone=HardNet682022.05 | 0.7459 | 0.826 | 0.8485 | 0.8652 | — | — | 0.8358 | — | — | — | — | |
| HarDNet-MSEGFPS=34.80, Parameters (Millions)=33.34, Flops (GMac)=6.022023.03 | 0.7459 | 0.826 | 0.8485 | 0.8652 | — | — | 0.8358 | — | — | — | — | |
| Swin-Unet2025.10 | 0.7438 | 0.853 | 0.8297 | 0.8778 | — | 0.9539 | — | — | — | — | — | |
| U-Net++Category=Train-based, Prompting Strategy=Prompt-free2026.03 | 0.7435 | 0.8219 | — | — | — | — | — | — | — | — | — | |
| U-Net++FPS=81.34, Parameters (Millions)=9.16, Flops (GMac)=34.652023.03 | 0.742 | 0.8228 | 0.8437 | 0.8607 | — | — | 0.8295 | — | — | — | — | |
| BCDU-Net2025.10 | 0.7404 | 0.8508 | 0.8074 | 0.8993 | — | 0.9543 | — | — | — | — | — | |
| SSFormer-LTrain Set=CVC-ClinicDB2022.03 | 0.7348 | 0.827 | — | — | — | — | — | — | — | — | — | |
| GPSParams. (M)=4.222023.12 | 0.725 | 0.881 | — | — | — | — | — | — | — | — | — | |
| SSFParams. (M)=4.262023.12 | 0.717 | 0.873 | — | — | — | — | — | — | — | — | — | |
| EffiSegNet-B7Pre-trained on external dataset=true2024.07 | 0.7073 | 0.7629 | 0.7713 | 0.8957 | — | — | — | — | — | 0.8289 | — | |
| ColonSegNet2022.05 | 0.698 | 0.792 | 0.8193 | 0.8432 | — | — | 0.7999 | — | — | — | — | |
| ColonSegNetFPS=73.95, Parameters (Millions)=5.01, Flops (GMac)=62.162023.03 | 0.698 | 0.792 | 0.8193 | 0.8432 | — | — | 0.7999 | — | — | — | — | |
| SSFormer-STrain Set=CVC-ClinicDB2022.03 | 0.6977 | 0.779 | — | — | — | — | — | — | — | — | — | |
| BiasParams. (M)=4.162023.12 | 0.691 | 0.865 | — | — | — | — | — | — | — | — | — | |
| MedSAMCategory=SAM-based, Prompting Strategy=Box prompts2026.03 | 0.6774 | 0.7921 | — | — | — | — | — | — | — | — | — |