ReNeg: Learning Negative Embedding with Reward Guidance
This work addresses the need for better negative embeddings in text-to-image generation applications, offering a method that generalizes across models, but it is incremental as it builds on existing negative embedding and reward feedback techniques.
The paper tackled the problem of suboptimal negative embeddings in text-to-image generation by introducing ReNeg, an end-to-end method that learns improved negative embeddings using reward guidance, resulting in significant improvements in human preference alignment over existing methods.
In text-to-image (T2I) generation applications, negative embeddings have proven to be a simple yet effective approach for enhancing generation quality. Typically, these negative embeddings are derived from user-defined negative prompts, which, while being functional, are not necessarily optimal. In this paper, we introduce ReNeg, an end-to-end method designed to learn improved Negative embeddings guided by a Reward model. We employ a reward feedback learning framework and integrate classifier-free guidance (CFG) into the training process, which was previously utilized only during inference, thus enabling the effective learning of negative embeddings. We also propose two strategies for learning both global and per-sample negative embeddings. Extensive experiments show that the learned negative embedding significantly outperforms null-text and handcrafted counterparts, achieving substantial improvements in human preference alignment. Additionally, the negative embedding learned within the same text embedding space exhibits strong generalization capabilities. For example, using the same CLIP text encoder, the negative embedding learned on SD1.5 can be seamlessly transferred to text-to-image or even text-to-video models such as ControlNet, ZeroScope, and VideoCrafter2, resulting in consistent performance improvements across the board.