Dimension-Reduction Attack! Video Generative Models are Experts on Controllable Image Synthesis

TL;DR

This paper introduces a novel method called DRA-Ctrl that repurposes video generative models for controllable image synthesis, leveraging their ability to model dynamic scenes for high-quality, lower-dimensional image tasks.

Contribution

The paper proposes a new paradigm for video-to-image knowledge transfer, including a mixup transition strategy and a tailored attention mechanism, enabling video models to excel in image generation tasks.

Findings

Video models outperform image-trained models in controllable image synthesis.

The proposed method achieves smooth transition from video to image generation.

Repurposed video models demonstrate untapped potential for diverse visual tasks.

Abstract

Video generative models can be regarded as world simulators due to their ability to capture dynamic, continuous changes inherent in real-world environments. These models integrate high-dimensional information across visual, temporal, spatial, and causal dimensions, enabling predictions of subjects in various status. A natural and valuable research direction is to explore whether a fully trained video generative model in high-dimensional space can effectively support lower-dimensional tasks such as controllable image generation. In this work, we propose a paradigm for video-to-image knowledge compression and task adaptation, termed \textit{Dimension-Reduction Attack} (\texttt{DRA-Ctrl}), which utilizes the strengths of video models, including long-range context modeling and flatten full-attention, to perform various generation tasks. Specially, to address the challenging gap between continuous video frames and discrete image generation, we introduce a mixup-based transition strategy that ensures smooth adaptation. Moreover, we redesign the attention structure with a tailored masking mechanism to better align text prompts with image-level control. Experiments across diverse image generation tasks, such as subject-driven and spatially conditioned generation, show that repurposed video models outperform those trained directly on images. These results highlight the untapped potential of large-scale video generators for broader visual applications. \texttt{DRA-Ctrl} provides new insights into reusing resource-intensive video models and lays foundation for future unified generative models across visual modalities. The project page is https://dra-ctrl-2025.github.io/DRA-Ctrl/.