2D Object Detection on LVIS v1.0
Metric: AP novel-Unrestricted open-vocabulary training (higher is better)
Metric: AP novel-Unrestricted open-vocabulary training (higher is better)
| # | Model↕ | AP novel-Unrestricted open-vocabulary training▼ | Augmentations | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | DITO | 45.8 | No | Region-centric Image-Language Pretraining for Op... | 2023-09-29 | Code |
| 2 | OWL-ViT (CLIP-L/14) | 31.2 | Yes | Simple Open-Vocabulary Object Detection with Vis... | 2022-05-12 | Code |
| 3 | ViLD-ensemble w/ ALIGN (Eb7-FPN) | 27 | No | Open-vocabulary Object Detection via Vision and ... | 2021-04-28 | Code |
| 4 | X-Paste | 22.8 | No | X-Paste: Revisiting Scalable Copy-Paste for Inst... | 2022-12-07 | Code |
| 5 | ViLD-ensemble (R152-FPN) | 19.8 | No | Open-vocabulary Object Detection via Vision and ... | 2021-04-28 | Code |
| 6 | ViLD-ensemble (R50-FPN) | 16.7 | No | Open-vocabulary Object Detection via Vision and ... | 2021-04-28 | Code |
| 7 | ViLD (R50-FPN) | 16.3 | No | Open-vocabulary Object Detection via Vision and ... | 2021-04-28 | Code |