ControlVideo: Training-free Controllable Text-to-Video Generation

TLDR

Text‑driven diffusion models have revolutionized image generation, yet video diffusion lags because of high training costs and problems such as appearance inconsistency and structural flicker, especially in long videos. The authors propose ControlVideo, a training‑free framework that enables natural and efficient text‑to‑video generation. ControlVideo, adapted from ControlNet, uses input motion sequences for coarse structural consistency and adds three modules—cross‑frame self‑attention for appearance coherence, an interleaved‑frame smoother with interpolation to reduce flicker, and a hierarchical sampler that stitches short clips into long videos. ControlVideo outperforms state‑of‑the‑arts on extensive motion‑prompt pairs and can generate short and long videos in minutes on a single NVIDIA 2080Ti. Code is available at https://github.com/YBYBZhang/ControlVideo.

Abstract

Text-driven diffusion models have unlocked unprecedented abilities in image generation, whereas their video counterpart still lags behind due to the excessive training cost of temporal modeling. Besides the training burden, the generated videos also suffer from appearance inconsistency and structural flickers, especially in long video synthesis. To address these challenges, we design a \emph{training-free} framework called \textbf{ControlVideo} to enable natural and efficient text-to-video generation. ControlVideo, adapted from ControlNet, leverages coarsely structural consistency from input motion sequences, and introduces three modules to improve video generation. Firstly, to ensure appearance coherence between frames, ControlVideo adds fully cross-frame interaction in self-attention modules. Secondly, to mitigate the flicker effect, it introduces an interleaved-frame smoother that employs frame interpolation on alternated frames. Finally, to produce long videos efficiently, it utilizes a hierarchical sampler that separately synthesizes each short clip with holistic coherency. Empowered with these modules, ControlVideo outperforms the state-of-the-arts on extensive motion-prompt pairs quantitatively and qualitatively. Notably, thanks to the efficient designs, it generates both short and long videos within several minutes using one NVIDIA 2080Ti. Code is available at https://github.com/YBYBZhang/ControlVideo.