Dynamic Hand Gesture Recognition

Gestures were most likely utilised by our ancestors to communicate. Armstrong once stated that he believes movements using the hands were the earliest form of complex human communication. The beginning stage of human computer Interactions is a gesture recognition system. Here, we have designed a Dynamic Hand Gesture Recognition (HGR) System using a Neural Network, which can recognize the gesture using the Computer or Laptop’s Web Camera and do the corresponding tasks. The model has been divided into mainly three modules. Firstly, Module 1 is about building a CNN model which we use to predict the gestures. Our second module is about predicting the gesture through live video feed. The final module is about assigning a specified task for a particular predicted gesture. For the predicted gesture, we use PyAutoGUI module to control devices like mouse and keyboard. Finally, the system is made to do some desired tasks in response to the gestures and obtains 99.84% of accuracy.