Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1675 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Point Cloud Classification | ModelNet40 (official split) | 20 | Feb 26, 2026 | ||
| Action Recognition | Epic-100 (test) | 20 | Feb 26, 2026 | ||
| Image-to-Text Retrieval | Crisscrossed Captions (CxC) | 20 | Jun 1, 2026 | ||
| Image Classification | NATS-Bench SSS ImageNet-16-120 (test) | 20 | Feb 26, 2026 | ||
| Image Classification | NATS-Bench SSS CIFAR-100 (val) | 20 | Feb 26, 2026 | ||
| Visual Question Answering | GQA OOD (test) | 20 | May 6, 2026 | ||
| Reinforcement Learning | Atari 100k steps (test) | 20 | Feb 26, 2026 | ||
| Point Cloud Upsampling | Point Cloud Upsampling (test) | 20 | Feb 26, 2026 | ||
| Action Recognition | Diving-48 (val) | 20 | Feb 26, 2026 | ||
| Temporal Action Localization | MultiTHUMOS | 20 | Feb 26, 2026 | ||
| Dialogue Relation Extraction | DialogRE v2 (test) | 20 | Feb 26, 2026 | ||
| Zero-shot Object Detection | MS-COCO (65/15) | 20 | Feb 26, 2026 | ||
| Few-shot Image Classification | mini-Plantae (test) | 20 | Feb 26, 2026 | ||
| Few-shot Image Classification | mini-Places (test) | 20 | Feb 26, 2026 | ||
| Few-shot Image Classification | mini-CUB (test) | 20 | Feb 26, 2026 | ||
| Underwater Image Enhancement | R90 (test) | 20 | Feb 26, 2026 | ||
| Underwater Image Enhancement | S1000 (test) | 20 | Feb 26, 2026 | ||
| Phrase Grounding | Flickr30k | 20 | Feb 26, 2026 | ||
| Classification | ModelNet40 original (test) | 20 | Feb 26, 2026 | ||
| Shadow Removal | ISTD original (test) | 20 | Feb 26, 2026 | ||
| Image Harmonization | iHarmony4 HCOCO | 20 | Feb 26, 2026 | ||
| Image Classification | CIFAR-100 90% noise (test) | 20 | Feb 26, 2026 | ||
| Image Classification | CIFAR-100 80% noise (test) | 20 | Feb 26, 2026 | ||
| Image Classification | CIFAR-100 50% noise (test) | 20 | Feb 26, 2026 | ||
| Image Classification | CIFAR-100 20% noise (test) | 20 | Feb 26, 2026 |