AIRMOUSE: AI-Powered Gesture Control

AIRMOUSE: AI-Powered Gesture Control

1Srinivas Gadige, 2Sanda Rohan,3Jamalpur Uday Kumar, 4N. Santhoshi

 1,2,3UG Student, 4Assistant Professor 1,2,3,4CSE- Artificial Intelligence and Machine Learning 1,2,3,4Sreenidhi Institute of Science and Technology, Hyderabad, Telangana.

Abstract : AirMouse is a novel, touchless cursor control system that leverages real-time hand- tracking and gesture-recognition technologies to replace conventional pointing devices. By combining OpenCV’s robust computer-vision pipelines with MediaPipe’s precise landmark detection and PyAutoGUI’s seamless interface for cursor actions, AirMouse translates intuitive hand movements— such as finger pinches, palm orientations, and swipe gestures—into standard mouse operations (click, drag, scroll, and cursor movement) without requiring physical contact. This approach not only enhances user comfort during prolonged use (e.g., reducing wrist strain and repetitive stress), but also significantly improves accessibility for individuals with motor disabilities and hygienic considerations in shared work environments. Preliminary evaluations demonstrate reliable performance across varying lighting conditions, achieving average tracking accuracy above 90% and click-detection latency under 150 ms on midrange hardware. Future work will focus on expanding the gesture vocabulary (multi-finger recognition), integrating adaptive learning to personalize gesture sensitivity, and exploring AR overlays for richer visual feedback.

Keywords: Touchless Interaction, Hand-Tracking, Gesture Recognition, OpenCV, MediaPipe, PyAutoGUI, Accessibility, Human–Computer Interaction