Loading the SOTA2 catalog…
SOTA2 Research · benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
| Task | Dataset | Domain | Trend | Results | Last Update |
|---|---|---|---|---|---|
| Classification | DVS-Gesture | 34 | May 25, 2026 | ||
| EEG emotion recognition | SEED Subject-independent | 34 | Mar 13, 2026 | ||
| Asymmetric Traveling Salesperson Problem | ATSP N=100 (test) | 34 | Jun 2, 2026 | ||
| Reinforcement Learning | HalfCheetah v3 | 34 | Apr 7, 2026 | ||
| 5-way 1-shot Classification | Omniglot (test) | 34 | Feb 26, 2026 | ||
| Grayscale Image Denoising |
|---|
| Urban100 (test) |
| 34 |
| Feb 26, 2026 |
| Person Re-identification | Market-1501 MA | 34 | Jun 16, 2026 |
|---|
| Human Pose and Shape Estimation | 3DPW 14 (test) | 34 | Feb 26, 2026 |
|---|
| Causal Discovery | Cancer | 34 | May 27, 2026 |
|---|
| Semantic Segmentation | DSEC (test) | 34 | Mar 25, 2026 |
|---|
| Image Classification | ImageNet-C level 3 (test) | 34 | Feb 26, 2026 |
|---|
| Semantic Segmentation | Pascal VOC Original protocol 732 labeled images | 34 | Feb 26, 2026 |
|---|
| Semantic Segmentation | Pascal VOC 366 labeled images (Original protocol) | 34 | Feb 26, 2026 |
|---|
| Semantic Segmentation | Pascal VOC 183 labeled images (Original protocol) | 34 | Feb 26, 2026 |
|---|
| Face Verification | LFW, AgeDB, CALFW, CPLFW, CFP-FP (10-fold cross-val) | 34 | Feb 26, 2026 |
|---|
| Medical Image Segmentation | TN3K | 34 | Jul 2, 2026 |
|---|
| Multimodal Machine Unlearning Evaluation | MLLMU-Bench (test) | 34 | Jul 9, 2026 |
|---|
| Autonomous Driving | Bench2Drive | 34 | May 4, 2026 |
|---|
| Conditional Image Retrieval | CTRATE-IR (test) | 34 | Feb 27, 2026 |
|---|
| MRI prostate segmentation | Prostate MRI (test) | 34 | Feb 26, 2026 |
|---|
| Object Detection | Foggy Cityscapes CF (test) | 34 | Feb 26, 2026 |
|---|
| Zero-shot Evaluation | Benchmarks 0-shot | 33 | Jul 9, 2026 |
|---|
| Causal Discovery | CausalMan Small | 33 | Jul 3, 2026 |
|---|
| Node Classification | Twitch Gamers Domain Skew Setting (Ratio 1:10:1:1:1:1) 30 Clients | 33 | Jul 2, 2026 |
|---|
| Composed Retrieval | Quilt-VQA Red | 33 | Jul 2, 2026 |
|---|