STIV: Scalable Text and Image Conditioned Video Generation

TL;DR

STIV introduces a scalable, unified framework for text and image conditioned video generation using diffusion transformers, achieving state-of-the-art results across multiple tasks with a simple design.

Contribution

The paper presents a systematic study and a simple, scalable method for text-image conditioned video generation, integrating image and text conditioning into a diffusion transformer architecture.

Findings

Achieves 83.1 on VBench T2V, surpassing existing models.

Achieves 90.1 on VBench I2V, setting a new state-of-the-art.

Demonstrates versatility across tasks like video prediction and frame interpolation.

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.