Panoptic Segmentation on Cityscapes open-vocabulary
39.6PQOpenSD
Evaluation results
| Method | Links | |
|---|---|---|
| OpenSDTraining Data=COCO, Backbone=Swin-L2023.12 | 39.6 | |
| OpenSDTraining Data=COCO+O365, Backbone=Swin-T2023.12 | 39 | |
| OpenSDTraining Data=COCO+O365, Backbone=R502023.12 | 38.6 | |
| OpenSDTraining Data=COCO, Backbone=Swin-T2023.12 | 38.2 | |
| X-DecoderTraining Data=COCO+ITP, Backbone=Swin-L2023.12 | 38.1 | |
| OPSNetTraining Data=COCO+IN, Backbone=R502023.12 | 37.8 | |
| OpenSDTraining Data=COCO, Backbone=R502023.12 | 37.8 | |
| OpenSeeDTraining Data=COCO, Backbone=Swin-L2023.12 | 37.7 | |
| OpenSeeDTraining Data=COCO+O365, Backbone=Swin-T2023.12 | 37.3 | |
| X-DecoderTraining Data=COCO+ITP, Backbone=Swin-T2023.12 | 37.2 | |
| OpenSeeDTraining Data=COCO, Backbone=Swin-T2023.12 | 35.7 |