Video Panoptic Segmentation On Cityscapes Vps
Métriques
VPQ
VPQ (stuff)
VPQ (thing)
Résultats
Résultats de performance de divers modèles sur ce benchmark
Tableau comparatif
| Nom du modèle | VPQ | VPQ (stuff) | VPQ (thing) |
|---|---|---|---|
| tarvis-a-unified-approach-for-target-based | 58.0 | 69.0 | 42.9 |
| tarvis-a-unified-approach-for-target-based | 53.3 | 66.0 | 35.9 |
| polyphonicformer-unified-query-learning-for | 62.3 | - | - |
| video-k-net-a-simple-strong-and-unified | 62.2 | 71.8 | 49.8 |
| vip-deeplab-learning-visual-perception-with | 63.1 | 73.0 | 49.5 |
| tarvis-a-unified-approach-for-target-based | 58.9 | 69.9 | 43.7 |
| learning-to-associate-every-segment-for-video | 57.3 | 66.4 | 44.7 |
| video-panoptic-segmentation-1 | 57.0 | 66.0 | 44.7 |