A Human-Centric Interactive Avatar Creation Framework Using Motion Capture and Face Expression Detection in Metaverse

초록

The metaverse is a next-generation internet that merges the physical and virtual worlds, offering users immersive interaction experiences. Avatars are essential components in the metaverse, as they are human digital representations that interact and communicate with virtual objects. However, current avatar development has limited movement capabilities (standing, walking, running, and jumping), and the features of facial expressions are limited. This paper presents an interactive avatar generation framework by leveraging motion capture and facial expression to enable human-like interactive avatars in metaverse ecosystems. The proposed framework utilizes the camera to capture the human skeleton data and provide more detailed human movements with facial expressions. This framework achieved high accuracy and low latency for detailed avatar movement synchronization in physical and virtual worlds.

키워드