Gesture-Based Computing: Revolutionizing Human-Computer Interaction

Abstract: Gesture-based computing (GBC) could be useful in volume control of many digital devices. It is a pivotal facet of human-computer interaction that can be tremendously useful for aging individuals or for ones with limited mobility. Utilizing computer vision techniques in conjunction with the Mediapipe hand landmark model is developed here. Such system empower users to adjust device volume through intuitive hand gestures. The proposed system tracks the movement of the thumb and index finger to gauge the inter-finger distance, subsequently regulating volume levels. To evaluate the accuracy of this method, we have computed the percentage of frames where hand gestures accurately translate to the desired volume adjustment. The outcomes of our analysis are depicted graphically, offering a lucid portrayal of volume control dynamics over time. This visual representation facilitates a comprehensive assessment of system performance and identifies avenues for enhancement. Moreover, we delve into the transformative potential of GBC technology across diverse domains like healthcare, education, automotive, and gaming, by delivering intuitive and contactless control mechanisms. Nevertheless, we acknowledge the obstacles encountered in ensuring consistent accuracy and adaptability across varying environments. By incorporating machine learning algorithms and rigorously refining sensor placement and calibration techniques, this research contributes to the continual advancement of GBC technology. The insights and recommendations provided in this paper are intended to assist researchers, developers, and industry stakeholders in their efforts to achieve smooth and efficient human-computer interaction. Our proposed work is novel as it incorporates adaptive normalization of finger distance relative to hand scale and lighting-adjusted calibration. This ensures consistent performance across users and environments, enhancing robustness and user inclusivity unlike existing systems relying solely on static distance thresholds.