Model card
APE-Large.
Tsinghua / MEGVIIopen-sourceUnknown paramsAligned vision encoder + region-text alignment with EVA-02 ViT-L backbone
Aligning and Prompting Everything All at Once for Universal Visual Perception. Achieves 66.4 mask AP on LVIS v1.0 minival with open-vocabulary capabilities. CVPR 2024. arXiv:2312.02153.
§ 01 · Benchmarks
Every benchmark APE-Large has a recorded score for.
| # | Benchmark | Area · Task | Metric | Value | Rank | Date | Source |
|---|---|---|---|---|---|---|---|
| 01 | LVIS v1.0 | Computer Vision · Object Detection | mask-ap-rare | 65.4% | #1 | 2023-12-04 | source ↗ |
| 02 | LVIS v1.0 | Computer Vision · Object Detection | box-ap | 70.3% | #2 | 2023-12-04 | source ↗ |
| 03 | LVIS v1.0 | Computer Vision · Object Detection | mask-ap | 66.4% | #2 | 2023-12-04 | source ↗ |
Rank column shows this model’s position vs all other models scored on the same benchmark + metric (competitors after the slash). #1 in red means current SOTA. Sorted by rank, then newest result.
§ 03 · Papers
1 paper with results for APE-Large.
- 2023-12-04· Computer Vision· 3 results
APE: Aligning and Prompting Everything All at Once for Universal Visual Perception
§ 05 · Sources & freshness
Where these numbers come from.
arxiv
3
results
3 of 3 rows marked verified.