Guest Editors
Dr. Stanisław Hożyń
Email: s.hozyn@amw.gdynia.pl
Affiliation: Faculty of Mechanics and Electrical Engineering, Polish Naval Academy, Gdynia, Poland
Homepage:
Research Interests: unmanned underwater vehicles, computer vision
Summary
Gestures offer an intuitive, contactless way to interact, supporting natural communication and enhancing safety in human–robot collaboration across various settings. As autonomous systems are deployed in factories, public spaces, and maritime or field operations, gesture recognition systems must perform reliably under real-world conditions, including changing viewpoints, occlusion, rapid motion, and complex backgrounds. Therefore, 3D and temporal computer vision methods such as depth or stereo sensing, hand–body pose estimation, and spatiotemporal modelling are essential for robust gesture recognition.
This Special Issue emphasises vision-centric gesture recognition for human–robot interaction and interactive autonomy. Submissions are sought on topics including 3D hand and body perception, continuous gesture recognition, and gesture-driven decision-making or control. In addition to computer vision, the integration of auxiliary modalities such as IMUs, radars, and wearables is encouraged, as these approaches enhance robustness and safety through uncertainty-aware recognition and runtime monitoring. Articles that connect perception to downstream action—planning, shared control, or verification—are particularly encouraged.
Suggested Themes
· 3D hand pose, mesh reconstruction, and fine-grained articulation modelling
· Body pose, skeleton-based representations, and multi-person interaction scenes
· Spatiotemporal architectures for the continuous recognition of gestures
· Gesture understanding for HRI: intent recognition, interaction events, command languages
· Gesture-conditioned robot/vehicle control, shared autonomy, and safety constraints
· Multimodal enhancements to vision (IMU, radar, wearables)
· Robustness against occlusion, motion blur, low light, adverse weather and domain adaptation
· Uncertainty estimation, confidence calibration, failure detection, and runtime monitoring
· Real-time and edge deployment, latency–accuracy trade-offs, calibration and synchronisation
Keywords
gesture recognition, human–robot interaction (HRI), spatiotemporal deep learning, multimodal sensor fusion, 3D hand pose estimation
Login
Register
Submit a Paper
Propose a Special lssue