CanonFuse3D: Human-Centric Canonical Fusion for Calibration-Free Dual-View 3D Pose Estimation in Skiing Videos
Published in MMSports 2026 (ACM Multimedia Workshop), 2026
Status: Accepted for oral presentation at the 2026 International ACM Workshop on Multimedia Content Analysis in Sports (MMSports), co-located with ACM Multimedia 2026.
This work examines human-centric canonicalization and dual-view 3D pose fusion for outdoor skiing videos. The aim is to combine independently estimated poses from moving cameras without relying on calibrated camera geometry at the fusion stage.
Links: Paper on OpenReview · ACM Multimedia 2026 technical program
The workshop presentation is upcoming as of October 2026; this entry does not imply the presentation has already taken place.
Recommended citation: Chen K, Luo N, Kimura T, Naito E, Xie C, Kitahara I. CanonFuse3D: Human-Centric Canonical Fusion for Calibration-Free Dual-View 3D Pose Estimation in Skiing Videos. MMSports 2026, ACM Multimedia Workshop (accepted, 2026).
Download Paper