STIV: Scalable Text and Image Conditioned Video Generation
Abstract
The field of video generation has made remarkable advancements, yet there remains a pressing need for a clear, systematic recipe that can guide the development of robust and scalable models. In this work, we present a comprehensive study that systematically explores the interplay of model architectures, training recipes, and data curation strategies, culminating in a simple and scalable text-image-conditioned video generation method, named STIV. Our framework integrates image condition into a Diffusion Transformer (DiT) through frame replacement, while incorporating text conditioning via a joint image-text conditional classifier-free guidance. This design enables STIV to perform both text-to-video (T2V) and text-image-to-video (TI2V) tasks simultaneously. Additionally, STIV can be easily extended to various applications, such as video prediction, frame interpolation, multi-view generation, and long video generation, etc. With comprehensive ablation studies on T2I, T2V, and TI2V, STIV demonstrate strong performance, despite its simple design. An 8.7B model with 512 resolution achieves 83.1 on VBench T2V, surpassing both leading open and closed-source models like CogVideoX-5B, Pika, Kling, and Gen-3. The same-sized model also achieves a state-of-the-art result of 90.1 on VBench I2V task at 512 resolution. By providing a transparent and extensible recipe for building cutting-edge video generation models, we aim to empower future research and accelerate progress toward more versatile and reliable video generation solutions.
Cite
@article{arxiv.2412.07730,
title = {STIV: Scalable Text and Image Conditioned Video Generation},
author = {Zongyu Lin and Wei Liu and Chen Chen and Jiasen Lu and Wenze Hu and Tsu-Jui Fu and Jesse Allardice and Zhengfeng Lai and Liangchen Song and Bowen Zhang and Cha Chen and Yiran Fei and Lezhi Li and Yizhou Sun and Kai-Wei Chang and Yinfei Yang},
journal= {arXiv preprint arXiv:2412.07730},
year = {2025}
}