Semantic Segmentation on Pascal Context 459 classes (test)
25.8mIoUCAT-Seg
Evaluation Results
| Method | Links | |
|---|---|---|
| CAT-SegVLM=SILC-C-L/162023.10 | 25.8 | |
| CAT-SegVLM=SILC-C-B/162023.10 | 22 | |
| CAT-SegVLM=SILC-S-B/162023.10 | 21.9 | |
| CAT-SegVLM=CLIP-G/142023.10 | 21.4 | |
| CAT-SegVLM=CLIP-L/142023.10 | 20.4 | |
| CAT-SegVLM=CLIP-B/162023.10 | 16.6 | |
| FreeSeg + MAFTOpen-vocabulary setting=true, Proposal Generator Backbone=ResNet101, CLIP Backbone=ViT-B/16, Module=IP-CLIP Encoder2023.09 | 12.8 | |
| OVSegVLM=CLIP-L/142023.10 | 12.4 | |
| OpenSegOpen-vocabulary setting=true, Additional training data=true2023.09 | 12.2 | |
| OVSegOpen-vocabulary setting=true2023.09 | 11 | |
| OVSegVLM=CLIP-B/162023.10 | 11 | |
| ZegFormerVLM=CLIP-B/162023.10 | 10.4 | |
| ZSsegVLM=CLIP-L/142023.10 | 10.2 | |
| FreeSegOpen-vocabulary setting=true, Proposal Generator Backbone=ResNet101, CLIP Backbone=ViT-B/162023.09 | 6.4 | |
| LSeg+Open-vocabulary setting=true2023.09 | 5.2 |