Training a Neural Network on Virtual Reality Devices: Challenges and Limitations

The processing power of Virtual Reality (VR) devices is constantly growing. However, few applications still take advantage of these capabilities. Machine learning algorithms have shown promise in enabling an immersive and personalized experience for VR device users. Therefore, it is interesting that these algorithms are processed directly on the devices themselves, without needing other external resources. In this work, a Neural Network (NN) is trained for real-time image classification using different VR devices. The results show the feasibility of incorporating VR devices for NN training without compromising the quality of the interaction, simply and saving external resources.