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 |
|---|---|---|---|---|---|
| Image-to-text Retrieval | MS-COCO (test) | 127 | Apr 28, 2026 | ||
| Algorithm Selection for Continuous Black-Box Optimisation | COCO BBOB Evaluation Protocols (LIO, Random, LPO) | 125 | Apr 13, 2026 | ||
| Image Reconstruction | COCO 2017 (val) | 123 | Mar 26, 2026 | ||
| Instance Segmentation | MS COCO | 123 | Jul 7, 2026 | ||
| Object Detection |
| MS COCO novel classes 2017 (val) |
| 123 |
| Feb 27, 2026 |
| Visual Hallucination Evaluation | MSCOCO | 122 | Jun 24, 2026 |
|---|
| Object Hallucination Evaluation | MS-COCO POPE Random | 121 | Jun 1, 2026 |
|---|
| Referring Image Segmentation | RefCOCO+ (testA) | 121 | Jun 23, 2026 |
|---|
| Image-to-Text Retrieval | MS-COCO 1K (test) | 121 | Feb 26, 2026 |
|---|
| Image Captioning | MS COCO (test) | 120 | Mar 16, 2026 |
|---|
| Semantic Segmentation | COCO | 119 | Jun 16, 2026 |
|---|
| Text-to-Image Retrieval | MSCOCO 1K (test) | 118 | Mar 31, 2026 |
|---|
| Object Hallucination Evaluation | POPE MSCOCO | 114 | Jun 30, 2026 |
|---|
| Referring Image Segmentation | RefCOCOg (val) | 114 | May 14, 2026 |
|---|
| Visual Question Answering | COCO | 106 | Apr 15, 2026 |
|---|
| Text-to-image generation | COCO | 104 | Jul 2, 2026 |
|---|
| Multi-Person Pose Estimation | COCO (test-dev) | 101 | Feb 26, 2026 |
|---|
| Semantic Segmentation | COCO Object (val) | 101 | May 7, 2026 |
|---|
| Multi-person Pose Estimation | COCO 2017 (test-dev) | 99 | Feb 26, 2026 |
|---|
| Watermark extraction | COCO | 98 | Feb 26, 2026 |
|---|
| Image Retrieval | MS-COCO (test) | 98 | Feb 26, 2026 |
|---|
| Few-shot Semantic Segmentation | COCO 5-shot 20i | 98 | Jun 26, 2026 |
|---|
| Referring Segmentation | refCOCO (val) | 96 | Jul 8, 2026 |
|---|
| Image-to-Text Retrieval | MSCOCO (1K test) | 96 | Mar 31, 2026 |
|---|
| Referring Segmentation | refCOCO (testA) | 95 | Jul 8, 2026 |
|---|