Search papers, labs, and topics across Lattice.
This paper introduces RTR-DiT, a real-time streaming video stylization framework based on a Diffusion Transformer (DiT). They fine-tune a bidirectional DiT model on a video stylization dataset and distill it into a few-step autoregressive model using Self Forcing and Distribution Matching Distillation. A novel reference-preserving KV cache update strategy enables stable long video processing and real-time style switching.
Achieve real-time, consistent video stylization with Diffusion Transformers, enabling interactive style switching and long video processing previously unattainable.
Recent advances in video generation models has significantly accelerated video generation and related downstream tasks. Among these, video stylization holds important research value in areas such as immersive applications and artistic creation, attracting widespread attention. However, existing diffusion-based video stylization methods struggle to maintain stability and consistency when processing long videos, and their high computational cost and multi-step denoising make them difficult to apply in practical scenarios. In this work, we propose RTR-DiT (DiT as Real-Time Rerenderer), a steaming video stylization framework built upon Diffusion Transformer. We first fine-tune a bidirectional teacher model on a curated video stylization dataset, supporting both text-guided and reference-guided video stylization tasks, and subsequently distill it into a few-step autoregressive model via post-training with Self Forcing and Distribution Matching Distillation. Furthermore, we propose a reference-preserving KV cache update strategy that not only enables stable and consistent processing of long videos, but also supports real-time switching between text prompts and reference images. Experimental results show that RTR-DiT outperforms existing methods in both text-guided and reference-guided video stylization tasks, in terms of quantitative metrics and visual quality, and demonstrates excellent performance in real-time long video stylization and interactive style-switching applications.
