HPD
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Text-to-image generation | HPD v2 (test) | ImageReward166.49 | 53 | |
| Text-to-Image Generation | HPD | PickScore64 | 38 | |
| Human Preference Evaluation | HPD v2 (test) | Preference Accuracy85.4 | 32 | |
| Human Preference Prediction | HPD v2 | Accuracy85.4 | 25 | |
| Human Preference Prediction | HPD v3 | Accuracy78.3 | 21 | |
| Text-to-Image Generation | HPD v1.0 (test) | HPS v30.155 | 19 | |
| Pairwise Preference | HPD v3 (test) | Accuracy86.7 | 18 | |
| Text-to-Image Preference Prediction | HPD v3 | Accuracy77.2 | 17 | |
| Text-to-Image Generation | HPD v3 | Characters Score11.91 | 15 | |
| Preference Prediction | HPD v3 (test) | Accuracy76.9 | 14 | |
| Text-to-Image Preference Evaluation | HPD v2 (test) | Aesthetic (Mean)6.182 | 14 | |
| Generative Model Evaluation | HPD v3 | Real Score171 | 13 | |
| Text-to-image preference evaluation | HPD v3 (test) | Accuracy78.3 | 11 | |
| Text-to-image generation evaluation | HPD 3200 prompts v2 | PickScore23.19 | 10 | |
| Human Preference Alignment | HPD v2 | HPS-v2.10.379 | 10 | |
| Text-to-Image Alignment | HPD v2.1 (test) | HPS-v2.10.374 | 10 | |
| Pairwise Preference Prediction | HPD v3 | Accuracy77.6 | 9 | |
| Pairwise Preference Prediction | HPD v2 | Accuracy85.6 | 9 | |
| Pairwise preference prediction | HPD v3++ v1 (test) | Aesthetic Score79.1 | 7 | |
| Text-to-Image Generation | HPD prompts | HPS v20.3265 | 7 | |
| Text-to-Image Generation | HPD v2 (val) | PickScore0.2213 | 7 | |
| Preference Alignment | HPD 2.1 (test) | HPSv315.236 | 7 | |
| Human Preference Alignment | HPD v2 (test) | HPSv2.10.371 | 7 | |
| Text-to-Image Preference Alignment | HPD v2.1 | HPS-v2.1 Score0.385 | 6 | |
| Text-to-Image Generation | HPD v2 | HPS v2.129.17 | 6 |
Showing 25 of 41 rows