Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 796 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Aspect Sentiment Triplet Extraction | 14Lap ASTE-DATA-V2 (test) | 32 | Feb 26, 2026 | ||
| Aspect Sentiment Triplet Extraction | 14Rest ASTE-DATA-V2 (test) | 32 | Feb 26, 2026 | ||
| Classification | banknote | 32 | May 13, 2026 | ||
| Anomaly Localization | MVTec (test) | 32 | Feb 26, 2026 | ||
| Anti-spoofing | ASVspoof LA 2019 (test) | 32 | Feb 26, 2026 | ||
| Pixel-level Manipulation Detection | DEFACTO 12k | 32 | Feb 27, 2026 | ||
| Semi-supervised Image Classification | ImageNet (10% labeled) | 32 | Feb 26, 2026 | ||
| Temporal Action Localization | ActivityNet 1.2 | 32 | Feb 27, 2026 | ||
| Chinese Word Segmentation | PKU (test) | 32 | Feb 26, 2026 | ||
| Continual Semantic Segmentation | Pascal-VOC 15-1 scenario 2012 | 32 | Feb 26, 2026 | ||
| Image Generation | Stacked MNIST | 32 | Feb 26, 2026 | ||
| Traffic flow forecasting | PeMS04 | 32 | Apr 21, 2026 | ||
| Disease progression prediction | Dutch Data Warehouse (DDW) March 2021 (test) | 32 | Feb 26, 2026 | ||
| Unsupervised document hashing | TMC | 32 | Feb 26, 2026 | ||
| Unsupervised document hashing | Reuters | 32 | Feb 26, 2026 | ||
| Targeted Hashing Attack | MS-COCO | 32 | Feb 26, 2026 | ||
| Targeted Hashing Attack | FLICKR-25K | 32 | Feb 26, 2026 | ||
| Human-Object Interaction Detection | V-COCO Scenario 1 1.0 | 32 | Mar 11, 2026 | ||
| Aspect-based Sentiment Classification | 15Rest SemEval-2015 (test) | 32 | Apr 16, 2026 | ||
| Universal Domain Adaptation | VisDA | 32 | Feb 26, 2026 | ||
| Slot filling | Restaurant-8K | 32 | Feb 26, 2026 | ||
| 3D Object Detection | Waymo Open Dataset 1.2 (val) | 32 | Feb 26, 2026 | ||
| Video Salient Object Detection | DAVSOD (test) | 32 | Feb 26, 2026 | ||
| RST Discourse Parsing | RST-DT Parseval (test) | 32 | Feb 26, 2026 | ||
| Machine Translation | newstest 2010-2016 (test) | 32 | Feb 26, 2026 |