Action Detection On Charades
Métriques
mAP
Résultats
Résultats de performance de divers modèles sur ce benchmark
Tableau comparatif
| Nom du modèle | mAP |
|---|---|
| pdan-pyramid-dilated-attention-network-for | 26.5 |
| r-c3d-region-convolutional-3d-network-for | 12.4 |
| pat-position-aware-transformer-for-dense | 26.5 |
| temporal-gaussian-mixture-layer-for-videos | 22.3 |
| avid-dataset-anonymized-videos-from-diverse | 23.2 |
| token-turing-machines | 28.79 |
| modeling-multi-label-action-dependencies-for | 23.7 |
| learning-latent-super-events-to-detect | 19.41 |
| coarse-fine-networks-for-temporal-activity | 25.1 |
| asynchronous-temporal-fields-for-action | 9.6 |
| self-supervised-pretraining-with | 26.95 |
| unimd-towards-unifying-moment-retrieval-and | 26.53 |
| avid-dataset-anonymized-videos-from-diverse | 25.2 |
| ctrn-class-temporal-relational-network-for | 27.8 |
| ms-tct-multi-scale-temporal-convtransformer | 25.4 |
| neural-message-passing-on-hybrid-spatio | 23.7 |