Semantic Segmentation on COCO
65.2mIoU Class (Mean)ODISE
Evaluation Results
| Method | Links | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ODISETraining Dataset=COCO, Supervision: label=true, Supervision: mask=true, Supervision: caption=false2023.03 | 65.2 | — | — | — | — | — | — | — | — | — | — | — | — | |
| LSeg+Training Dataset=COCO, Supervision: label=true, Supervision: mask=true, Supervision: caption=false2023.03 | 55.1 | — | — | — | — | — | — | — | — | — | — | — | — | |
| ODISETraining Dataset=COCO, Supervision: label=false, Supervision: mask=false, Supervision: caption=true2023.03 | 52.4 | — | — | — | — | — | — | — | — | — | — | — | — | |
| PFENetBackbone=ResNet-101, Shots=5-Shot, Label Resizing=Fixed (473)2020.08 | 42.7 | 40.4 | 46.8 | 43.2 | 40.5 | 52.3 | 70 | 69.5 | 71.3 | 65.8 | — | — | — | |
| PFENetBackbone=VGG-16, Shots=5-Shot, Label Resizing=Fixed (473)2020.08 | 40.4 | 38.2 | 42.5 | 41.8 | 38.9 | 53.5 | 68.3 | 68.2 | 70.1 | 65 | — | — | — | |
| PFENetBackbone=ResNet-101, Shots=1-Shot, Label Resizing=Fixed (473)2020.08 | 38.5 | 36.8 | 41.8 | 38.7 | 36.7 | 51.6 | 65.9 | 66.6 | 66 | 63 | — | — | — | |
| PFENetBackbone=VGG-16, Shots=5-Shot, Label Resizing=Original2020.08 | 37.7 | 35.9 | 40.7 | 38.1 | 36.1 | 50.3 | 65.2 | 65.2 | 65.5 | 61.6 | — | — | — | |
| PFENetBackbone=ResNet-101, Shots=5-Shot, Label Resizing=Original2020.08 | 37.4 | 38.5 | 38.6 | 38.2 | 34.3 | 51.5 | 65.6 | 65.7 | 64.7 | 61.9 | — | — | — | |
| PFENetBackbone=VGG-16, Shots=1-Shot, Label Resizing=Fixed (473)2020.08 | 36.3 | 35.4 | 38.1 | 36.8 | 34.7 | 53.3 | 66.1 | 66.6 | 67.1 | 63.3 | — | — | — | |
| OpenSegTraining Dataset=COCO, Supervision: label=false, Supervision: mask=false, Supervision: caption=true2023.03 | 36.1 | — | — | — | — | — | — | — | — | — | — | — | — | |
| PFENetBackbone=VGG-16, Shots=1-Shot, Label Resizing=Original2020.08 | 34.1 | 33.4 | 36 | 34.1 | 32.8 | 50 | 63.1 | 63.5 | 63.4 | 60 | — | — | — | |
| PFENetBackbone=ResNet-101, Shots=1-Shot, Label Resizing=Original2020.08 | 32.4 | 34.3 | 33 | 32.3 | 30.1 | 52.2 | 59.5 | 61.5 | 61.4 | 58.6 | — | — | — | |
| PANet2019Backbone=VGG-16, Shots=5-Shot, Label Resizing=Fixed (473)2020.08 | 29.7 | — | — | — | — | — | — | — | — | 63.5 | — | — | — | |
| FWBF2019Backbone=ResNet-101, Shots=5-Shot, Label Resizing=Original2020.08 | 23.7 | 19.1 | 21.5 | 23.9 | 30.1 | — | — | — | — | — | — | — | — | |
| FWBF2019Backbone=VGG-16, Shots=5-Shot, Label Resizing=Original2020.08 | 22.6 | 20.9 | 19.2 | 21.9 | 28.4 | — | — | — | — | — | — | — | — | |
| FWBF2019Backbone=ResNet-101, Shots=1-Shot, Label Resizing=Original2020.08 | 21.2 | 19.9 | 18 | 21 | 28.9 | — | — | — | — | — | — | — | — | |
| GroupViTTraining Dataset=GCC+YFCC, Supervision: label=false, Supervision: mask=false, Supervision: caption=true2023.03 | 21.1 | — | — | — | — | — | — | — | — | — | — | — | — | |
| PANet2019Backbone=VGG-16, Shots=1-Shot, Label Resizing=Fixed (473)2020.08 | 20.9 | — | — | — | — | — | — | — | — | 59.2 | — | — | — | |
| FWBF2019Backbone=VGG-16, Shots=1-Shot, Label Resizing=Original2020.08 | 20 | 18.4 | 16.7 | 19.6 | 25.4 | — | — | — | — | — | — | — | — | |
| A-MCG2019Backbone=ResNet-101, Shots=1-Shot, Label Resizing=Original2020.08 | — | — | — | — | — | — | — | — | — | 52 | — | — | — | |
| A-MCG2019Backbone=ResNet-101, Shots=5-Shot, Label Resizing=Original2020.08 | — | — | — | — | — | — | — | — | — | 54.7 | — | — | — | |
| CaGNetTraining Mode=Open Vocabulary2023.03 | — | — | — | — | — | — | — | — | — | — | 35.6 | 13.4 | 19.5 | |
| FreeSegTraining Mode=Full Supervision2023.03 | — | — | — | — | — | — | — | — | — | — | 42.9 | 54.3 | 47.9 | |
| FreeSegTraining Mode=Open Vocabulary2023.03 | — | — | — | — | — | — | — | — | — | — | 42.2 | 49.1 | 45.3 | |
| SPNetTraining Mode=Open Vocabulary2023.03 | — | — | — | — | — | — | — | — | — | — | 34.6 | 26.9 | 30.3 | |
| STRICTTraining Mode=Open Vocabulary2023.03 | — | — | — | — | — | — | — | — | — | — | 35.3 | 30.3 | 32.6 | |
| ZegFormerTraining Mode=Open Vocabulary2023.03 | — | — | — | — | — | — | — | — | — | — | 36.6 | 33.2 | 34.8 | |
| ZS5Training Mode=Open Vocabulary2023.03 | — | — | — | — | — | — | — | — | — | — | 34.9 | 10.6 | 16.2 | |
| ZSSegTraining Mode=Open Vocabulary2023.03 | — | — | — | — | — | — | — | — | — | — | 39.6 | 43.6 | 41.5 |