HunyuanCustom: A Multimodal-Driven Architecture for Customized Video Generation
Research paper by Teng Hu and collaborators. Published 2025-05-07. Topics: multi-modal, video generation, identity consistency, image-text fusion, LLaVA.
Resource information
- Authors
- Teng Hu, Zhentao Yu, Zhengguang Zhou, Sen Liang, Yuan Zhou, Qin Lin, Qinglin Lu
- Research topics
- multi-modal, video generation, identity consistency, image-text fusion, LLaVA, image ID enhancement, temporal concatenation, AudioNet, hierarchical alignment, spatial cross-attention, video-driven, latent-compressed
- Publication date
- 2025-05-07
- Publication
- arXiv
- Last record update
- Nov 3, 2026