BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Asia/Hong_Kong
X-LIC-LOCATION:Asia/Hong_Kong
BEGIN:STANDARD
TZOFFSETFROM:+0800
TZOFFSETTO:+0800
TZNAME:HKT
DTSTART:19911015T033000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20251218T030657Z
LOCATION:Meeting Room S421\, Level 4
DTSTART;TZID=Asia/Hong_Kong:20251217T153400
DTEND;TZID=Asia/Hong_Kong:20251217T154500
UID:siggraphasia_SIGGRAPH Asia 2025_sess140_papers_2307@linklings.com
SUMMARY:MODepth: Benchmarking Mobile Multi-frame Monocular Depth Estimatio
 n with Optical Image Stabilization
DESCRIPTION:Yu Lu (Shanghai Jiao Tong University), Hao Pan (Microsoft Rese
 arch Asia), Dian Ding and Jiatong Ding (Shanghai Jiao Tong University), Yo
 ngjian Fu (Central South University), Yi-Chao Chen (Shanghai Jiao Tong Uni
 versity), Ju Ren (Tsinghua University), and Guangtao Xue (Shanghai Jiao To
 ng University)\n\nThis paper presents MODepth, a multi-frame monocular dep
 th estimation system based on the controlled motion of an optical image st
 abilization (OIS) module. By actively injecting acoustic signals, we induc
 e regular translational movements of the OIS lens, resulting in controllab
 le camera pose changes and simplifying inter-frame pose estimation. Levera
 ging multi-frame images captured under OIS-controlled lens movements, we d
 esign a high-precision depth estimation network, MODNet, and introduce the
  principal point offset estimation module and pose estimation modules to f
 ully exploit geometric information across frames. To validate the effectiv
 eness of our approach, we collect a new dataset MODdata with 1100 samples 
 in nearly 220 indoor scenarios and benchmark our model as an OIS-based mul
 ti-frame depth estimation method, comparing it to ground truth obtained fr
 om a depth sensor and other state-of-the-art monocular depth estimation al
 gorithms.  Our method achieves competitive or superior performance compare
 d to fully supervised baselines, reaching an RMSE of 0.439, which outperfo
 rms all evaluated methods, demonstrating that self-supervised fine-tuning 
 with OIS-induced parallax is a viable alternative to ground-truth supervis
 ion.\n\nRegistration Category: Full Access, Full Access Supporter\n\nSessi
 on Chair: Haisen Zhao (Shandong University)\n\n
END:VEVENT
END:VCALENDAR