This study proposes a dual-stream framework integrating RGB and thermal sensors for UAV swarms conducting autonomous mapping of dense forests, ensuring high-resolution coverage and real-time adaptation to canopy variations.

Harold Vincent McDougal, Li Wenjie, Sahana Kripa Balachandran, Thierry Claude Dumont, Kim Minho, Anastasia Gabrielle Russo

This paper introduces a visual scene understanding framework using attention-guided multi-scale feature aggregation, enhancing underwater robots’ capabilities in detecting objects and terrain under turbid and low-light conditions.

Gerard Raymond Foster, Yu Lianhua, Sharanya Maheshwari, Petra Louise Anderson, Zhang Qingsheng, Omar Farouk Ismail

This study develops a real-time gesture prediction model enabling seamless sign language interpretation in AR systems, enhancing accessibility and interaction through motion-captured datasets and predictive temporal modeling.

Edward Nathaniel Klein, Lu Zhiqiang, Aarti Shyam Bhandari, Sergio Paolo Vitale, Amira Yasmin Khalil, Sun Taeyang

This work presents a depth-aware visual odometry system using sparse light field cameras, improving pose estimation and trajectory tracking in narrow corridor navigation tasks for indoor autonomous agents.

Samuel Eric Donovan, Zhang Liwei, Priya Anushka Pillai, Jerome Walter Fischer, Chen Jinglong, Olivia Jane Cartwright

This paper develops low-power vision systems optimized for extended autonomous surveillance in remote regions, employing adaptive frame-rate modulation and hardware-accelerated inference to prolong operational life without sacrificing detection accuracy.

Nathan Charles Berkley, Li Xiaohan, Sumitha Vignesh Karthik, Sebastian Emil Rausch, Lin Mei, Lucia Irene Gatti