Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1775 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Image Classification | CIFAR-10 60% symmetric noise (test) | 19 | Feb 26, 2026 | ||
| Image Classification | CIFAR-10 40% symmetric noise (test) | 19 | Feb 26, 2026 | ||
| Video Classification | Moments in Time v1 (val) | 19 | Feb 26, 2026 | ||
| Camouflaged Object Segmentation | CHAMELEON 76 images (test) | 19 | Feb 26, 2026 | ||
| Text-Image Retrieval | Fashion-Gen | 19 | Apr 30, 2026 | ||
| Object Detection | MS-COCO (48/17) | 19 | Feb 27, 2026 | ||
| Semantic Segmentation | GTAV (val) | 19 | Feb 26, 2026 | ||
| 3D Object Detection | KITTI Cars (IoU3D ≥ 0.7) (test) | 19 | Feb 26, 2026 | ||
| Monocular Depth Estimation | DIW | 19 | Feb 26, 2026 | ||
| Action Segmentation | 50Salads mid granularity | 19 | Feb 27, 2026 | ||
| Text-to-Image Generation | Multi-modal CelebA-HQ | 19 | Feb 26, 2026 | ||
| Image Deraining | Rain800 Test100 (test) | 19 | Feb 26, 2026 | ||
| Node Classification | Pubmed v1 (test) | 19 | Feb 26, 2026 | ||
| Panoptic Segmentation | SYNTHIA to Cityscapes | 19 | Feb 26, 2026 | ||
| Continual Learning | Split-MNIST | 19 | Feb 26, 2026 | ||
| Relation Extraction | CDR (dev) | 19 | Feb 26, 2026 | ||
| Temporal Action Localization (Verb) | EPIC-Kitchens 100 (val) | 19 | Feb 26, 2026 | ||
| Trajectory Forecasting | TrajNet++ real world Type III (test) | 19 | Feb 26, 2026 | ||
| Anomaly Detection | CIFAR-10-C | 19 | Feb 26, 2026 | ||
| Classification | CIFAR-10 10% IDN (test) | 19 | Feb 26, 2026 | ||
| Object Detection | MS-COCO 10% labeled protocol | 19 | Feb 26, 2026 | ||
| Generalized Zero-Shot Retrieval (Text-to-Audio) | AudioSet ZSL (test) | 19 | Feb 26, 2026 | ||
| Node classification | ACM (20% train) | 19 | Apr 20, 2026 | ||
| Image Classification | PASCAL VOC 2007 | 19 | Feb 26, 2026 | ||
| Named Entity Recognition | CoNLL NER (test) | 19 | Feb 26, 2026 |