Awesome RL with Human Feedback / 2023 / arXiv

Better Aligning Text-to-Image Models with Human Preference

作者：Xiaoshi Wu, Keqiang Sun, Feng Zhu, Rui Zhao, Hongsheng Li · 单位：Multimedia Laboratory, The Chinese University of Hong Kong, Qing Yuan Research Institute, Shanghai Jiao Tong University, Centre for Perceptual and Interactive Intelligence (CPII), Shanghai AI Laboratory · 来源：arXiv · 发布日期：2023

偏好学习对齐目标设计多模态反馈文生图对齐
新标签打开 AlphaXiv新标签打开 Code
Better Aligning Text-to-Image Models with Human Preference figure
AlphaXiv 中文概览（可滚动查看）