AI-Powered Virtual Mouse: Enhancing Real-Time Gesture Control with Latency-Aware Jitter Suppression

In recent years, the need for interaction between humans and computers without any physical devices like keyboards or a mouse has led to many innovations. This research paper explores the hand gesture-based mouse control system that allows the user to perform mouse operations by doing certain hand gestures with the help of a webcam, which captures the hand gesture and transfers the gesture to the system, which gives the specific response based on the gesture by performing the mouse operations. The system mainly focuses on the sensitivity of the control and the cooldown mechanisms to ensure stable operation and prevent accidental triggers. Testing shows that users easily adapted to the system and learned the basic controls within 2–5 minutes. The system maintains consistent performance at 30 frames per second (FPS), with gesture recognition times averaging 110–123 ms, contributing to smooth cursor control and responsive gesture recognition. Click accuracy reached 98.7%, and gesture misinterpretation was reduced to 1.3%, ensuring the reliable execution of operations. This system provides humans with a more comfortable experience on mouse control, and it also helps challenged persons to control the mouse efficiently.