MISTA: compact mutil-identity structure-aware tensorized avatars

Key Points

• Enhanced avatar representations utilize multi-identity structures to optimize user interaction.
• The framework introduces compact data representation for efficient processing of avatar identities.
• Methodological integration of tensorized avatars facilitates dynamic identity mapping across platforms.
• These findings may enable advanced applications in gaming, virtual reality, and online identity management.

Abstract