Human Image Animation
Human image animation focuses on generating realistic videos of humans from a single image or video clip, often driven by user-specified poses or actions. Current research heavily utilizes diffusion models, often enhanced with 3D parametric models like SMPL, to improve temporal consistency, address challenges like identity preservation and motion blur, and enable finer control over both human and camera movement. This field is significant for its potential applications in film production, virtual try-ons, and other creative industries, while also raising important ethical considerations regarding potential misuse of generated content.
Papers
DisPose: Disentangling Pose Guidance for Controllable Human Image Animation
Hongxiang Li, Yaowei Li, Yuhang Yang, Junjie Cao, Zhihong Zhu, Xuxin Cheng, Long Chen
Enhancing Facial Consistency in Conditional Video Generation via Facial Landmark Transformation
Lianrui Mu, Xingze Zhou, Wenjie Zheng, Jiangnan Ye, Xiaoyu Liang, Yuchen Yang, Jianhong Bai, Jiedong Zhuang, Haoji Hu