Publications
カンファレンス (国際) Motion Context-Based Adaptation for Voice Navigation in Indoor Exploration
Shohei Yoshinari (Ritsumeikan University), Kota Tsubouchi, Hirotaka Noda (Ritsumeikan University), Yuuki Nishiyama (The University of Tokyo), Masamichi Shimosaka (Institute of Science Tokyo), Nobuhiko Nishio (Ritsumeikan University)
The 16th International Conference on Indoor Positioning and Indoor Navigation (IPIN 2026)
2026.11.18
Voice navigation using earable devices can support users during indoor exploration, but voice notifications are difficult to ignore once they are presented. Therefore, inappropriate notification timing may interrupt users when they are moving, looking around, or focusing on objects in dynamic indoor environments. This paper proposes a two-stage framework for voice notification timing decision that estimates motion contexts, which are intermediate representations defined by combinations of body motion and head motion, from Inertial Measurement Unit (IMU) data collected by an earable device and a smartphone, and then determines whether a voice notification should be allowed or suppressed. Experiments in a real bookstore environment showed that the proposed framework improved notification acceptance and suppression of unacceptable notifications compared with fixed notification delivery and direct estimation from raw IMU features. These results indicate that motion contexts are an effective intermediate representation for adaptive voice notification timing in indoor exploration.
Paper :
Motion Context-Based Adaptation for Voice Navigation in Indoor Exploration
(外部サイト)