Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
May 16, 2026Frontiers in Virtual RealityOpen Access

Act-Env matching: activity and context-aware avatar positioning for telepresence in heterogeneous environments

View Full Paper
Ask AI
Bookmark
Share

Authors

HDHideki DeguchiSTShun TaguchiTHTakumi Hamazaki

Discussion

Loading...

Member takes

Overview

Randomized trial demonstrates improved avatar placement in varied environments, suggesting increased user preference and experience.

Key Points

  • This research aims to enhance avatar positioning in augmented reality telepresence by using activity and context awareness.
  • Developed an activity- and context-aware avatar positioning method called Act-Env Matching using a multimodal large language model.
  • Extracted keyframes from the local environment video to determine appropriate avatar placement.
  • Back-projected pixel locations to 3D coordinates using depth maps.
  • Participants preferred the activity- and context-matched avatar placements over traditional methods.
  • The system successfully positioned avatars without requiring spatial similarity between local and remote environments.
  • Two experiments indicated that users perceived the placements as appropriate.

Cite This Study

Deguchi et al. (2026) studied this question.

synapsesocial.com/papers/6a0808ffa487c87a6a40b068https://doi.org/10.3389/frvir.2026.1791810
View Full Paper
Ask AI
Bookmark
Share