AirWave: Hands-Free Cursor Navigation with Face and Voice
G Thippeswamy, R Akash, Ankit Suresh Savalagi, M K Dayanidhi, R Dileep · 2025
Context: Hands-free cursor navigation is considered indispensable for improving accessibility for physically disabled individuals. Currently, the control of mouse actions is based on facial gestures and voice commands in most existing systems. However, these solutions often face constraints such as a high sensitivity to environments, user fatigue, and reliance on advance hardware. Therefore, the following discussion will describe a lightweight, yet scalable system optimized for real-world conditions. Objectives: The project AirWave develops a hands-free cursor navigation system through facial gestures and voice commands offline. This is focused on real-time performance while minimizing the requirements for hardware on the system, thus becoming accessible and user-friendly for people with disabilities. This application aims at relating facial gestures like head tilts and blinks to cursor actions and integrating voice commands for advanced controls. Method: The system uses Dlib for facial landmark detection, OpenCV for video processing, and Vosk for offline speech recognition. The system reads in real-time video input from a webcam, detects facial gestures, and maps these to cursor movements using PyAutoGUI. Voice commands predefined trigger mouse actions, such as opening applications or scrolling. Result: The study conducted on the research papers provided critical insights into the current advancements and limitations of hands-free cursor navigation systems. The findings highlight the need for larger datasets and more sophisticated models to improve the accuracy. Conclusion: AirWave demonstrates the potential for accessible hands-free computing by addressing environmental sensitivity and hardware constraints. It provides a scalable, efficient solution, with future scope in multilingual commands, adaptive recognition, and IoT integration