Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 421 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Trajectory Prediction | H3D | 48 | Feb 26, 2026 | ||
| Crowd Counting | UCF-QNRF | 48 | Feb 26, 2026 | ||
| Vision-Language Navigation | R2R unseen v1.0 (val) | 48 | May 27, 2026 | ||
| Image Reconstruction | Urban100 (val) | 48 | Feb 26, 2026 | ||
| Image Reconstruction | BSD100 (val) | 48 | Feb 26, 2026 | ||
| RGBD Saliency Detection | SSB | 48 | Feb 26, 2026 | ||
| Few-shot classification | Meta-Dataset (test) | 48 | Feb 26, 2026 | ||
| Log-likelihood estimation | MNIST dynamically binarized (test) | 48 | Feb 26, 2026 | ||
| Nuclear Instance Segmentation | CPM 17 | 48 | May 19, 2026 | ||
| Action Detection | AVA v2.1 (val) | 48 | Feb 26, 2026 | ||
| Scene Completion | NYU v2 (test) | 48 | Mar 27, 2026 | ||
| Gait Recognition | CASIA-B BG (Bag) (BG#1-2 probe) | 48 | Feb 27, 2026 | ||
| Semantic Segmentation | ADE20K | 48 | May 8, 2026 | ||
| Semantic Image Synthesis | Cityscapes (test) | 48 | May 18, 2026 | ||
| Inference Speed | Generic Synthetic Data | 48 | Feb 26, 2026 | ||
| Abstractive Summarization | CNN/DailyMail full length F-1 (test) | 48 | Feb 26, 2026 | ||
| Machine Reading Comprehension | SQuAD 1.1 (dev) | 48 | Feb 26, 2026 | ||
| 3D Object Detection | KITTI cars (val) | 48 | Feb 26, 2026 | ||
| Question Answering | SearchQA (test) | 48 | Feb 26, 2026 | ||
| Person Re-identification | Market-1501 Multi. Query 1.0 | 48 | Feb 26, 2026 | ||
| Language Modeling | Yahoo (test) | 48 | Feb 26, 2026 | ||
| Person Re-identification | PRID 2011 (test) | 48 | Feb 26, 2026 | ||
| Text-guided molecule generation | ChEBI-20 (test) | 48 | Feb 26, 2026 | ||
| Tensor Decomposition | Hyperspectral image tensor | 48 | Feb 26, 2026 | ||
| Tensor Reconstruction | COIL-100 1.0 (test) | 48 | Feb 26, 2026 |