Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 2259 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Text-to-Video Generation | VideoPhy2 (ALL) | 16 | Apr 10, 2026 | ||
| Image Retrieval | Stanford Cars 196 (test) | 16 | Feb 26, 2026 | ||
| Semantic Segmentation | TROSD | 16 | Feb 26, 2026 | ||
| Object Detection | NEU-DET (test) | 16 | Mar 17, 2026 | ||
| Hand-Object Interaction Reconstruction | ARCTIC (test) | 16 | Feb 26, 2026 | ||
| Instruction-Guided Video Editing | OpenVE-Bench 1.0 (full) | 16 | Feb 26, 2026 | ||
| Deepfake Detection | Universal Deepfake Detection Evaluation Suite | 16 | Feb 26, 2026 | ||
| Remote Sensing Image Restoration | RRSHID 57 | 16 | Feb 26, 2026 | ||
| Sandstorm Image Restoration | DID 4 | 16 | Feb 26, 2026 | ||
| Underwater Image Restoration | UIEB 23 | 16 | Feb 26, 2026 | ||
| Nighttime Haze Removal | NHRW 51 | 16 | Feb 26, 2026 | ||
| Daytime Haze Removal | FTD 7 | 16 | Feb 26, 2026 | ||
| Anatomical Structure Segmentation | Combined laparoscopic datasets (Dresden, CholecSeg8k, AutoLaparoT3, EndoScapes-CVS201, M2caiSeg) (test) | 16 | Feb 26, 2026 | ||
| Tissue Segmentation | Combined (Dresden, CholecSeg8k, AutoLaparoT3, EndoScapes-CVS201, M2caiSeg) (test) | 16 | Feb 26, 2026 | ||
| Laparoscopic Segmentation | Gynsurg (unseen) | 16 | Feb 26, 2026 | ||
| Surgical Instrument Segmentation | Surgical Instrument combined (test) | 16 | Feb 26, 2026 | ||
| Interpretable Brain Pattern Discovery | NSD Measured fMRI | 16 | Feb 26, 2026 | ||
| Image Signal Processing | S24 1.0 (test) | 16 | Feb 26, 2026 | ||
| Novel View Synthesis | RealEstate10K v1 (hold-out scenes) | 16 | Feb 26, 2026 | ||
| Novel View Synthesis | DL3DV v1 (hold-out scenes) | 16 | Feb 26, 2026 | ||
| Retinal OCT Classification | NEH (five-fold patient-level cross-validation) | 16 | Feb 26, 2026 | ||
| Text-centric Visual Question Answering (Open-Ended) | Text-centric VQA benchmarks (Clean) | 16 | Feb 26, 2026 | ||
| Object-centric Typographic Robustness (Binary Choice) | Typo-D Attack | 16 | Feb 26, 2026 | ||
| Object-centric Typographic Robustness (Binary Choice) | Typo-D (Clean) | 16 | Feb 26, 2026 | ||
| Referring Video Object Segmentation | MeViS v2 (val) | 16 | May 20, 2026 |