- Artificial Intelligence in Simulation-based Training
- Understanding and explaining the adversarial property of deep neural networks
To provide feedback to trainees in virtual training environments to support skill learning and decision making.
Virtual reality (VR) based training is a modern training methodology that involves practice and learning in simulated (as opposed to physical) environments such as VR surgery/flight/driving simulators. Providing real-time tutoring feedback in such an environment is a challenging task. In this project, we explore and apply supervised data mining techniques to generate real-time feedback from available expert/novice behavioural samples, and validate these techniques in a VR surgical environment.