Surgical Phase Recognition on Cholec80 (test)
86.89PrecisionSKiT + EGTP
Evaluation Results
| Method | Links | ||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| SKiT + EGTPBackbone=SKiT, TEC=Excluded, EGTP=Included2026.05 | 86.89 | — | — | — | 92.21 | 85.53 | 76.15 | — | — | — | — | — | — | — | — | — | 61.95 | 0.78 | |
| Surgformer + TEC + EGTPBackbone=Surgformer, TEC=Included, EGTP=Included2026.05 | 86.89 | — | — | — | 92.2 | 88.7 | 78.3 | — | — | — | — | — | — | — | — | — | 60.53 | 0.71 | |
| SKiT + TEC + EGTPBackbone=SKiT, TEC=Included, EGTP=Included2026.05 | 86.88 | — | — | — | 92.54 | 85.77 | 76.64 | — | — | — | — | — | — | — | — | — | 62.4 | 0.69 | |
| Surgformer + EGTPBackbone=Surgformer, TEC=Excluded, EGTP=Included2026.05 | 86.71 | — | — | — | 91.6 | 87.69 | 77.01 | — | — | — | — | — | — | — | — | — | 54.12 | 1.09 | |
| SKiTBackbone=SKiT, TEC=Excluded, EGTP=Excluded2026.05 | 86.42 | — | — | — | 91.91 | 86.7 | 76.38 | — | — | — | — | — | — | — | — | — | 19.45 | 4.59 | |
| Surgformer + TECBackbone=Surgformer, TEC=Included, EGTP=Excluded2026.05 | 86.39 | — | — | — | 91.78 | 89.33 | 78.45 | — | — | — | — | — | — | — | — | — | 16.6 | 5.54 | |
| SurgformerBackbone=Surgformer, TEC=Excluded, EGTP=Excluded2026.05 | 86.22 | — | — | — | 91.22 | 87.97 | 76.98 | — | — | — | — | — | — | — | — | — | 15.77 | 5.85 | |
| SKiT + TECBackbone=SKiT, TEC=Included, EGTP=Excluded2026.05 | 85.83 | — | — | — | 91.98 | 88.15 | 76.97 | — | — | — | — | — | — | — | — | — | 20.82 | 3.47 | |
| Trans-SVNet + EGTPBackbone=Trans-SVNet, TEC=Excluded, EGTP=Included2026.05 | 84.96 | — | — | — | 89.09 | 83.35 | 72.64 | — | — | — | — | — | — | — | — | — | 51.52 | 1.19 | |
| Trans-SVNetMulti-task learning=false, 10-second relaxed boundary metrics=false2023.05 | 84.7 | — | — | — | — | 83.6 | 72.5 | — | — | — | — | — | — | — | — | — | — | — | |
| Trans-SVNet + TEC + EGTPBackbone=Trans-SVNet, TEC=Included, EGTP=Included2026.05 | 84.34 | — | — | — | 89.13 | 83.53 | 72.56 | — | — | — | — | — | — | — | — | — | 54.97 | 1.16 | |
| Trans-SVNetBackbone=Trans-SVNet, TEC=Excluded, EGTP=Excluded2026.05 | 84.08 | — | — | — | 88.61 | 83.23 | 71.74 | — | — | — | — | — | — | — | — | — | 12.54 | 6.42 | |
| Trans-SVNet + TECBackbone=Trans-SVNet, TEC=Included, EGTP=Excluded2026.05 | 83.43 | — | — | — | 89.01 | 83.94 | 71.98 | — | — | — | — | — | — | — | — | — | 14.29 | 5.87 | |
| LoViTMulti-task learning=false, 10-second relaxed boundary metrics=false2023.05 | 83.1 | — | — | — | — | 86.5 | 74.2 | — | — | — | — | — | — | — | — | — | — | — | |
| ZENBackbone=ViT-B, Fine-tuning=true, Pre-training type=Existing Pretrained2026.02 | 82.03 | — | — | — | — | 86.9 | 72.67 | 81.2 | 90.04 | — | — | — | — | — | — | — | — | — | |
| MIS-DINOv2Backbone=ViT-L, Fine-tuning=true, Pre-training type=Self-supervised2026.02 | 81.46 | — | — | — | — | 86.22 | 71.69 | 80.65 | 89.61 | — | — | — | — | — | — | — | — | — | |
| MIS-MAEBackbone=ViT-B, Fine-tuning=true, Pre-training type=Self-supervised2026.02 | 79.24 | — | — | — | — | 82.49 | 68.02 | 77.32 | 86.8 | — | — | — | — | — | — | — | — | — | |
| MIS-DINOv2Backbone=ViT-B, Fine-tuning=true, Pre-training type=Self-supervised2026.02 | 79.24 | — | — | — | — | 85.26 | 70.05 | 79.01 | 87.74 | — | — | — | — | — | — | — | — | — | |
| SurgeNetBackbone=CAFormer18, Fine-tuning=true, Pre-training type=Existing Pretrained2026.02 | 79.22 | — | — | — | — | 83.53 | 68.22 | 77.42 | 87.46 | — | — | — | — | — | — | — | — | — | |
| MIS-DINOBackbone=ViT-B, Fine-tuning=true, Pre-training type=Self-supervised2026.02 | 79.21 | — | — | — | — | 86.47 | 70.4 | 79.39 | 87.56 | — | — | — | — | — | — | — | — | — | |
| PeskaVLPBackbone=ResNet50, Fine-tuning=true, Pre-training type=Existing Pretrained2026.02 | 78.52 | — | — | — | — | 81.95 | 66.11 | 76.22 | 85.64 | — | — | — | — | — | — | — | — | — | |
| MIS-MSNBackbone=ViT-B, Fine-tuning=true, Pre-training type=Self-supervised2026.02 | 77.45 | — | — | — | — | 80.58 | 66.03 | 74.55 | 85.66 | — | — | — | — | — | — | — | — | — | |
| AVTMulti-task learning=false, 10-second relaxed boundary metrics=false2023.05 | 77.3 | — | — | — | — | 82.1 | 66.4 | — | — | — | — | — | — | — | — | — | — | — | |
| MIS-MoCoV3Backbone=ViT-B, Fine-tuning=true, Pre-training type=Self-supervised2026.02 | 77.27 | — | — | — | — | 78.91 | 64.35 | 73.98 | 85.09 | — | — | — | — | — | — | — | — | — | |
| EndoViTBackbone=ViT-B, Fine-tuning=true, Pre-training type=Existing Pretrained2026.02 | 76.9 | — | — | — | — | 79.59 | 64.14 | 74.01 | 84.16 | — | — | — | — | — | — | — | — | — | |
| ImageNetBackbone=ViT-B, Fine-tuning=true, Pre-training type=Existing Pretrained2026.02 | 75.26 | — | — | — | — | 79.41 | 62.44 | 73.24 | 83.46 | — | — | — | — | — | — | — | — | — | |
| EndoFMBackbone=ViT-B, Fine-tuning=true, Pre-training type=Existing Pretrained2026.02 | 75.19 | — | — | — | — | 77.47 | 61.13 | 71.75 | 82.93 | — | — | — | — | — | — | — | — | — | |
| GSViTBackbone=EfficientViT, Fine-tuning=true, Pre-training type=Existing Pretrained2026.02 | 54.54 | — | — | — | — | 47.04 | 33.56 | 42.48 | 63.22 | — | — | — | — | — | — | — | — | — | |
| Baseline CLIPTraining Configuration=Baseline2026.03 | — | — | — | — | 3.3 | — | — | — | — | 0.4 | 2.8 | 0.53 | 0.32 | 3.14 | 0.58 | 50.61 | — | — | |
| Cholec80 FT (Init: JIGSAWS)Training Configuration=Cholec80 FT (Init: JIGSAWS)2026.03 | — | — | — | — | 70.25 | — | — | — | — | 65.63 | 75.61 | 51.16 | 74.24 | 61.3 | 44.65 | 65.45 | — | — | |
| Cholec80 Only FTTraining Configuration=Cholec80 Only FT2026.03 | — | — | — | — | 19.51 | — | — | — | — | 96.28 | 2.87 | 71.92 | 24.8 | 18.16 | 17.02 | 0 | — | — | |
| DINOArch=ViT-S, Pretrain=IN1K2023.08 | — | — | 77.4 | 72.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DINOArch=RN50, Pretrain=Pub2023.08 | — | — | — | 81.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Endo-FMVenue=MICCAI, Year=20232026.03 | — | — | — | — | 82.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| EndoFM-LVVenue=JBHI, Year=20252026.03 | — | — | — | — | 85.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| EndoMambaVenue=MICCAI, Year=20252026.03 | — | — | — | — | 84.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| EndoSSLVenue=MICCAI, Year=20232026.03 | — | — | — | — | 83 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FAMEVenue=CVPR, Year=20222026.03 | — | — | — | — | 81.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FPRLVenue=Ours, Year=-2026.03 | — | — | — | — | 85.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FSArch=RN50, Pretrain=IN1K2023.08 | — | — | — | 80.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| JIGSAWS FTTraining Configuration=JIGSAWS FT2026.03 | — | — | — | — | 12.23 | — | — | — | — | 77.89 | 17.26 | 1.9 | 0.58 | 0 | 0 | 2.92 | — | — | |
| M2CRLVenue=NeurIPS, Year=20242026.03 | — | — | — | — | 83.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MSNArch=ViT-S, Pretrain=Pub2023.08 | — | — | 83.4 | 80.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MSNArch=ViT-B, Pretrain=Pub2023.08 | — | — | 82.6 | 82.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MSNArch=ViT-L, Pretrain=Pub2023.08 | — | — | 84 | 82 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MSNArch=ViT-S, Pretrain=Pri2023.08 | — | — | 87 | 84.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MSNArch=ViT-B, Pretrain=Pri2023.08 | — | — | 87.3 | 85.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MSNArch=ViT-L, Pretrain=Pri2023.08 | — | — | 89.6 | 86.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| OperAArch=RN50, Pretrain=IN1K2023.08 | — | — | 84.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| OursSingle-Frame Finetuning=No2023.12 | — | 70.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| OursSingle-Frame Finetuning=Yes2023.12 | — | 79.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Ours + TCNSingle-Frame Finetuning=No2023.12 | — | 74.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Ours + TCNSingle-Frame Finetuning=Yes2023.12 | — | 81.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ProViCoVenue=CVPR, Year=20222026.03 | — | — | — | — | 82.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ST-AdapterVenue=NeurIPS, Year=20222026.03 | — | — | — | — | 81 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| STRGSingle-Frame Finetuning=Yes2023.12 | — | 77.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TeCNOSingle-Frame Finetuning=No2023.12 | — | 64.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TeCNOSingle-Frame Finetuning=Yes, Source=reported from [20]2023.12 | — | 80.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TeCNOArch=RN50, Pretrain=IN1K2023.08 | — | — | 83.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VideoMambaVenue=ECCV, Year=20242026.03 | — | — | — | — | 84.1 | — | — | — | — | — | — | — | — | — | — | — | — | — |