RefDrop: Controllable Consistency in Image or Video Generation via Reference Feature Guidance

TL;DR

RefDrop introduces a simple, controllable method for enhancing consistency in image and video generation by manipulating attention modules, eliminating the need for complex fine-tuning or additional encoders.

Contribution

The paper reveals that reference feature guidance operates as a linear interpolation of attention modules and proposes RefDrop, a rank-1 coefficient-free algorithm for precise control in diffusion models.

Findings

RefDrop achieves high consistency in multi-subject image generation.

It enables diverse content suppression and personalized video generation.

Outperforms state-of-the-art methods in controllability and quality.

Abstract

There is a rapidly growing interest in controlling consistency across multiple generated images using diffusion models. Among various methods, recent works have found that simply manipulating attention modules by concatenating features from multiple reference images provides an efficient approach to enhancing consistency without fine-tuning. Despite its popularity and success, few studies have elucidated the underlying mechanisms that contribute to its effectiveness. In this work, we reveal that the popular approach is a linear interpolation of image self-attention and cross-attention between synthesized content and reference features, with a constant rank-1 coefficient. Motivated by this observation, we find that a rank-1 coefficient is not necessary and simplifies the controllable generation mechanism. The resulting algorithm, which we coin as RefDrop, allows users to control the influence of reference context in a direct and precise manner. Besides further enhancing consistency in single-subject image generation, our method also enables more interesting applications, such as the consistent generation of multiple subjects, suppressing specific features to encourage more diverse content, and high-quality personalized video generation by boosting temporal consistency. Even compared with state-of-the-art image-prompt-based generators, such as IP-Adapter, RefDrop is competitive in terms of controllability and quality while avoiding the need to train a separate image encoder for feature injection from reference images, making it a versatile plug-and-play solution for any image or video diffusion model.