Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1090 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Open-domain Question Answering | TriviaQA (TQA) (test) | 26 | Feb 26, 2026 | ||
| Video highlight detection | TVsum | 26 | Feb 26, 2026 | ||
| Atlas Building | OAI knee MRI NIMH Data Archive Collection ID: 2343 (test) | 26 | Feb 26, 2026 | ||
| Incremental Semantic Segmentation | COCO | 26 | Jun 4, 2026 | ||
| Continuous Control | MuJoCo Ant | 26 | Mar 13, 2026 | ||
| Image Classification | Caltech-UCSD Birds (CUB-200-2011) (test) | 26 | Mar 4, 2026 | ||
| Intent Prediction | PhotoChat (test) | 26 | Feb 26, 2026 | ||
| Video Individual Counting | CroHD (test) | 26 | Feb 26, 2026 | ||
| Speech Enhancement | DNS Challenge Without Reverb (test) | 26 | Jul 3, 2026 | ||
| Super-Resolution | DIV2KRK (test) | 26 | Feb 26, 2026 | ||
| Named Entity Recognition | OntoNotes to I2B2, CoNLL, WNUT, GUM 5.0 (test) | 26 | Feb 26, 2026 | ||
| 2D Object Detection | SUN-RGBD (val) | 26 | Feb 26, 2026 | ||
| Lane Detection | LLAMAS (val) | 26 | Feb 26, 2026 | ||
| Hyperspectral Image Reconstruction | KAIST simulation (Average test) | 26 | Feb 26, 2026 | ||
| Event Extraction | ACE05 Evt | 26 | Feb 26, 2026 | ||
| Medical Image Segmentation | ACDC (5-fold cross-validation) | 26 | Feb 26, 2026 | ||
| Video Frame Interpolation | Xiph4K | 26 | Feb 26, 2026 | ||
| Video Person Re-identification | DukeMTMC-VideoReID | 26 | Feb 26, 2026 | ||
| Question Answering | WebSRC (dev) | 26 | Feb 26, 2026 | ||
| Link Prediction | ogbl-collab (val) | 26 | Feb 26, 2026 | ||
| Link Prediction | LastFM | 26 | Apr 20, 2026 | ||
| No-Reference Image Quality Assessment | LIVE 80/20 split | 26 | Feb 26, 2026 | ||
| Offline Reinforcement Learning | D4RL antmaze-med (diverse) | 26 | Feb 26, 2026 | ||
| 3D Point Cloud Registration | KITTI (test) | 26 | Feb 26, 2026 | ||
| Traffic Signal Control | New York real-world (test) | 26 | Feb 26, 2026 |