Object Detection On Ua Detrac
Métriques
mAP
Résultats
Résultats de performance de divers modèles sur ce benchmark
Tableau comparatif
| Nom du modèle | mAP |
|---|---|
| objects-as-points | 83.48 |
| faster-r-cnn-towards-real-time-object | 58.45 |
| ffavod-feature-fusion-architecture-for-video | 88.10 |
| yolo9000-better-faster-stronger | 57.72 |
| 3d-detnet-a-single-stage-video-based-vehicle | 53.30 |
| video-sparse-transformer-with-attention | 90.39 |
| r-fcn-object-detection-via-region-based-fully | 69.87 |
| spotnet-self-attention-multi-task-network-for | 86.8 |
| rn-vid-a-feature-fusion-architecture-for | 70.57 |