The field of character animation and behavior understanding is rapidly advancing, with a focus on developing more realistic and controllable animation methods. Recent research has explored the use of diffusion models, transformers, and other deep learning architectures to improve the quality and expressiveness of animated characters. One notable direction is the development of frameworks that can animate characters with dynamic backgrounds, allowing for more realistic and engaging videos. Another area of research is the use of motion transfer and retargeting techniques to enable the animation of characters with diverse identities, poses, and spatial configurations. Additionally, there is a growing interest in developing methods for behavior understanding, such as group activity detection and social behavior analysis, which can be applied to various fields like public safety, intelligent surveillance, and human-computer interaction. Noteworthy papers in this area include DiTalker, which proposes a unified DiT-based framework for speaking style-controllable portrait animation, and CharacterShot, which introduces a controllable and consistent 4D character animation framework. Other notable papers include X-UniMotion, which presents a unified and expressive implicit latent representation for whole-body human motion, and Animate-X++, which proposes a universal animation framework based on DiT for various character types.