[A-17] Biometrics

9:30 AM - 9:45 AM JST(12:30 AM - 12:45 AM UTC)

[A-17-01] Skeleton-based Signer Identification for Sign Language Communication Support

〇Manabu Okawa¹ (1. Tsukuba Univ. of Technology)

Keywords:

biometric authentication,signer identification,Dynamic Time Warping,template matching

This study presents a fundamental investigation of signer identification based on skeletal information for supporting sign language communication. Fourteen skeletal features of Pose and Hands were extracted from sign language videos using MediaPipe, and template matching based on dynamic time warping was applied. Experiments using a public dataset demonstrated clear visualization of individual differences and achieved an average identification accuracy of approximately 82%, confirming the effectiveness of the proposed approach.