DC-VideoGen: Efficient Video Generation with Deep Compression Video Autoencoder

Junyu Chen, Wenkun He, Yuchao Gu, Yuyang Zhao, Jincheng YU, Junsong Chen, Dongyun Zou, Yujun Lin, Zhekai Zhang, Muyang Li, Haocheng Xi, Ligeng Zhu, Enze Xie, Song Han, Han Cai

Keywords: efficient video diffusion model, deep compression video autoencoder, post-training model acceleration

Abstract: We introduce DC-VideoGen, a post-training acceleration framework for efficient video generation. DC-VideoGen can be applied to any pre-trained video diffusion model, improving efficiency by adapting it to a deep compression latent space with lightweight fine-tuning. The framework builds on two key innovations: (i) a **Deep Compression Video Autoencoder** with a novel chunk-causal temporal design that achieves 32x/64x spatial and 4x temporal compression while preserving reconstruction quality and generalization to longer videos; and (ii) **AE-Adapt-V**, a robust adaptation strategy that enables rapid and stable transfer of pre-trained models into the new latent space. Adapting the pre-trained Wan-2.1-14B model with DC-VideoGen requires only 10 GPU days on the NVIDIA H100 GPU. The accelerated models achieve up to 14.8x lower inference latency than their base counterparts without compromising quality, and further enable 2160x3840 video generation on a single GPU. We will release both the code and pre-trained models upon publication.

Primary Area: generative models

Submission Number: 3191