Gesture Recognition System in Virtual Reality

I created the system that can classify human gestures, which can provide essential feedback for human-AI interaction, in a virtual reality environment. I built the system which takes human motions as input using depth cameras, puts the input data in the physics engine and displays input data using VR equipment. I then implemented a system that can learn and classify human gestures using a model based on Long Short-Term Memory (LSTM) in real-time. After training the model, I successfully controlled an agent in the virtual reality environment by hand gestures.