CogVideoX: Text-to-Video Diffusion Models with An Expert Transformer

We present CogVideoX, a large-scale text-to-video generation model based on diffusion transformer, which can generate 10-second continuous videos aligned with text prompt, with a frame rate of 16 fps and resolution of 768 * 1360 pixels. Previous video generation models often had limited movement and...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2024-10
Main Authors: Yang, Zhuoyi, Teng, Jiayan, Zheng, Wendi, Ding, Ming, Huang, Shiyu, Xu, Jiazheng, Yang, Yuanming, Hong, Wenyi, Zhang, Xiaohan, Feng, Guanyu, Yin, Da, Gu, Xiaotao, Zhang, Yuxuan, Wang, Weihan, Cheng, Yean, Liu, Ting, Xu, Bin, Dong, Yuxiao, Tang, Jie
Format: Article
Language:English
Subjects:
Alignment
Data processing
Diffusion layers
Human performance
Performance evaluation
State-of-the-art reviews
Transformers
Video compression
Video data
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We present CogVideoX, a large-scale text-to-video generation model based on diffusion transformer, which can generate 10-second continuous videos aligned with text prompt, with a frame rate of 16 fps and resolution of 768 * 1360 pixels. Previous video generation models often had limited movement and short durations, and is difficult to generate videos with coherent narratives based on text. We propose several designs to address these issues. First, we propose a 3D Variational Autoencoder (VAE) to compress videos along both spatial and temporal dimensions, to improve both compression rate and video fidelity. Second, to improve the text-video alignment, we propose an expert transformer with the expert adaptive LayerNorm to facilitate the deep fusion between the two modalities. Third, by employing a progressive training and multi-resolution frame pack technique, CogVideoX is adept at producing coherent, long-duration, different shape videos characterized by significant motions. In addition, we develop an effective text-video data processing pipeline that includes various data preprocessing strategies and a video captioning method, greatly contributing to the generation quality and semantic alignment. Results show that CogVideoX demonstrates state-of-the-art performance across both multiple machine metrics and human evaluations. The model weight of both 3D Causal VAE, Video caption model and CogVideoX are publicly available at https://github.com/THUDM/CogVideo.
ISSN:2331-8422