A Comprehensive Survey on Human Video Generation: Challenges, Methods, and Insights
Wentao Lei, Jinting Wang, Fengji Ma, Guanjie Huang, Li Liu
TL;DR
This survey addresses the problem of generating realistic 2D human videos under controllable conditions such as text, audio, and pose. It classifies methods into text-driven, audio-driven, and pose-driven approaches, and analyzes datasets, evaluation metrics, and practical challenges, emphasizing the rise of diffusion-based video generation and its counterparts. Key contributions include a comprehensive taxonomy, a comparative review of modality-specific techniques, and a discussion of datasets, metrics, and commercial precedents, as well as guidance for choosing baselines and future directions. The findings highlight persistent challenges in appearance and pose continuity, background harmony, temporal synchronization, and computational efficiency, while identifying promising directions like large-scale data, long-form video, finer-grained control, and interactive systems with practical impact for entertainment and digital humans.
Abstract
Human video generation is a dynamic and rapidly evolving task that aims to synthesize 2D human body video sequences with generative models given control conditions such as text, audio, and pose. With the potential for wide-ranging applications in film, gaming, and virtual communication, the ability to generate natural and realistic human video is critical. Recent advancements in generative models have laid a solid foundation for the growing interest in this area. Despite the significant progress, the task of human video generation remains challenging due to the consistency of characters, the complexity of human motion, and difficulties in their relationship with the environment. This survey provides a comprehensive review of the current state of human video generation, marking, to the best of our knowledge, the first extensive literature review in this domain. We start with an introduction to the fundamentals of human video generation and the evolution of generative models that have facilitated the field's growth. We then examine the main methods employed for three key sub-tasks within human video generation: text-driven, audio-driven, and pose-driven motion generation. These areas are explored concerning the conditions that guide the generation process. Furthermore, we offer a collection of the most commonly utilized datasets and the evaluation metrics that are crucial in assessing the quality and realism of generated videos. The survey concludes with a discussion of the current challenges in the field and suggests possible directions for future research. The goal of this survey is to offer the research community a clear and holistic view of the advancements in human video generation, highlighting the milestones achieved and the challenges that lie ahead.