| Benchmark | Model | Rank | Results |
|---|---|---|---|
| phrase-grounding-on-flickr30k-entities-test | GLIPv2 | #1 | R@1: 87.7 |
| referring-expression-segmentation-on-phrasecut | GLIPv2 | #1 | Mean IoU: 61.3 |
| object-detection-on-odinw-full-shot-13-tasks | GLIPv2 | #4 | AP: 70.4 |
| object-detection-on-lvis-v1-0-minival | GLIPv2 | #5 | box AP: 59.8 |
| visual-question-answering-on-vqa-v2-test-std | GLIPv2-H (fine-tuned) | #17 | overall: 74.8 |
| object-detection-on-coco-test-dev | GLIPv2 (CoSwin-H, multi-scale) | #19 | box mAP: 62.4 |
| instance-segmentation-on-coco-test-dev | GLIPv2-H (fine-tuned) | #25 | mask AP: 48.9 |