InfinityStar: Unified Spacetime AutoRegressive Modeling for Visual Generation

TL;DR

InfinityStar is a novel unified autoregressive framework that efficiently generates high-resolution images and videos by modeling spatial and temporal dependencies within a single discrete architecture, outperforming existing methods.

Contribution

It introduces the first discrete autoregressive model capable of producing 720p videos at industrial quality and speed, unifying various generation tasks in a single framework.

Findings

Scores 83.74 on VBench, outperforming all autoregressive models.

Generates 5s, 720p videos 10x faster than diffusion methods.

First discrete autoregressive video generator at industrial resolution.

Abstract

We introduce InfinityStar, a unified spacetime autoregressive framework for high-resolution image and dynamic video synthesis. Building on the recent success of autoregressive modeling in both vision and language, our purely discrete approach jointly captures spatial and temporal dependencies within a single architecture. This unified design naturally supports a variety of generation tasks such as text-to-image, text-to-video, image-to-video, and long interactive video synthesis via straightforward temporal autoregression. Extensive experiments demonstrate that InfinityStar scores 83.74 on VBench, outperforming all autoregressive models by large margins, even surpassing some diffusion competitors like HunyuanVideo. Without extra optimizations, our model generates a 5s, 720p video approximately 10x faster than leading diffusion-based methods. To our knowledge, InfinityStar is the first discrete autoregressive video generator capable of producing industrial level 720p videos. We release all code and models to foster further research in efficient, high-quality video generation.