计算机科学
图像(数学)
集合(抽象数据类型)
生成模型
忠诚
生成语法
人工智能
功能(生物学)
机器学习
计算机视觉
模式识别(心理学)
电信
进化生物学
生物
程序设计语言
作者
Kimin Lee,Hao Liu,Moonkyung Ryu,Olivia Watkins,Yuqing Du,Craig Boutilier,Pieter Abbeel,Mohammad Ghavamzadeh,Shixiang Gu
出处
期刊:Cornell University - arXiv
日期:2023-01-01
被引量:28
标识
DOI:10.48550/arxiv.2302.12192
摘要
Deep generative models have shown impressive results in text-to-image synthesis. However, current text-to-image models often generate images that are inadequately aligned with text prompts. We propose a fine-tuning method for aligning such models using human feedback, comprising three stages. First, we collect human feedback assessing model output alignment from a set of diverse text prompts. We then use the human-labeled image-text dataset to train a reward function that predicts human feedback. Lastly, the text-to-image model is fine-tuned by maximizing reward-weighted likelihood to improve image-text alignment. Our method generates objects with specified colors, counts and backgrounds more accurately than the pre-trained model. We also analyze several design choices and find that careful investigations on such design choices are important in balancing the alignment-fidelity tradeoffs. Our results demonstrate the potential for learning from human feedback to significantly improve text-to-image models.
科研通智能强力驱动
Strongly Powered by AbleSci AI