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 |
|---|---|---|---|---|---|
| Phase classification | Pouring (test) | 36 | Feb 26, 2026 | ||
| Video Classification | YouTube-8M 2017 (val) | 36 | Feb 26, 2026 | ||
| 5-way Few-shot Classification | CUB (test) | 36 | Feb 26, 2026 | ||
| Periocular Recognition | VISOB Database (Experiment 1) | 36 | Feb 27, 2026 | ||
| Speech Synthesis | LJ Speech (test) | 36 | Feb 26, 2026 | ||
| Image Classification |
| CIFAR-10 Symmetry-50% noise (test) |
| 36 |
| Feb 26, 2026 |
| Trajectory Prediction | ZARA2 v1.0 (test) | 36 | Feb 26, 2026 |
|---|
| Entity Linking | MSNBC | 36 | Feb 26, 2026 |
|---|
| Visual Dialog | VisDial | 36 | Feb 26, 2026 |
|---|
| Action Recognition | N-UCLA | 36 | Feb 26, 2026 |
|---|
| Foreground-Background Segmentation | LSP (test) | 36 | Feb 26, 2026 |
|---|
| CS reconstruction | Set14 | 36 | Feb 26, 2026 |
|---|
| Extractive Summarization | CNN/Daily Mail (test) | 36 | Feb 26, 2026 |
|---|
| Multivariate Time Series Forecasting | Exchange-Rate (test) | 36 | Feb 26, 2026 |
|---|
| Multivariate Time Series Forecasting | Traffic (test) | 36 | May 26, 2026 |
|---|
| Monocular 3D Human Pose and Mesh Recovery | Human3.6M (test) | 36 | Feb 26, 2026 |
|---|
| Session-based recommendation | YOOCHOOSE (1/4) | 36 | Feb 27, 2026 |
|---|
| Crowd counting | UCSD crowd-counting (test) | 36 | Feb 26, 2026 |
|---|
| Video Recognition | Kinetics (val) | 36 | Feb 26, 2026 |
|---|
| Phrase Detection | VRD (test) | 36 | Feb 27, 2026 |
|---|
| Action Recognition | UTD-MHAD (cross-subject) | 36 | Feb 26, 2026 |
|---|
| Image Denoising | BSD (test) | 36 | Feb 26, 2026 |
|---|
| Machine Reading Comprehension | Daily Mail (val) | 36 | Feb 26, 2026 |
|---|
| Caption Retrieval | Flickr30k (test) | 36 | Feb 26, 2026 |
|---|
| Pedestrian Detection | Caltech (test) | 36 | Feb 26, 2026 |
|---|