Abstract
Generative artificial intelligence (AI) has made significant progress acrossvarious domains in recent years. Building on the rapid advancements in 2D,video, and 3D content generation fields, 4D generation has emerged as a noveland rapidly evolving research area, attracting growing attention. 4D generationfocuses on creating dynamic 3D assets with spatiotemporal consistency based onuser input, offering greater creative freedom and richer immersive experiences.This paper presents a comprehensive survey of the 4D generation field,systematically summarizing its core technologies, developmental trajectory, keychallenges, and practical applications, while also exploring potential futureresearch directions. The survey begins by introducing various fundamental 4Drepresentation models, followed by a review of 4D generation frameworks builtupon these representations and the key technologies that incorporate motion andgeometry priors into 4D assets. We summarize five major challenges of 4Dgeneration: consistency, controllability, diversity, efficiency, and fidelity,accompanied by an outline of existing solutions to address these issues. Wesystematically analyze applications of 4D generation, spanning dynamic objectgeneration, scene generation, digital human synthesis, 4D editing, andautonomous driving. Finally, we provide an in-depth discussion of the obstaclescurrently hindering the development of the 4D generation. This survey offers aclear and comprehensive overview of 4D generation, aiming to stimulate furtherexploration and innovation in this rapidly evolving field. Our code is publiclyavailable at: https://github.com/MiaoQiaowei/Awesome-4D.