WHOOPS
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Scene captioning | WHOOPS! (test) | CIDEr85.82 | 22 | |
| Scene captioning | WHOOPS! RGBP seen scenes (test) | CIDEr80.46 | 22 | |
| Compositional Visual Question Answering | WHOOPS! Compositional VQA | VQA BEM67.8 | 14 | |
| Text-to-Image Alignment | Whoops | VQAScore80.2 | 12 | |
| Identification of weird images | WHOOPS | Accuracy80 | 9 | |
| Image Captioning | WHOOPS (test) | BLEU@431 | 8 | |
| Text-to-Image Generation | Whoops | CLIP Score0.296 | 4 | |
| Text-to-Image Generation | Whoops | HPSv2 Score0.293 | 2 |
Showing 8 of 8 rows