Utilizing Unity and Machine Learning for Virtual Reality
DGIST Research Internship
Developed an intelligent VR simulation using Unity (C#) and LSTM models for realistic surgical training, enabling motion data acquisition, prediction, and real-time feedback.
Project Overview
During my research internship at the Intelligent Bio Opto Mechatronics Lab, DGIST, I worked on a project that integrates Unity and Machine Learning to develop intelligent Virtual Reality (VR) systems for surgical training. The project focused on creating a simulation environment in Unity with C#, mimicking microsurgical tools and hardware setups such as VR controllers and linear motors. This environment enabled synthetic data acquisition of positional and motion signals, which I processed and used to train Long Short-Term Memory (LSTM) models. These models captured temporal patterns in tool movements, enabling accurate prediction and enhancing the realism of VR-based microsurgery training. The work highlighted how VR combined with ML can provide cost-effective, immersive, and risk-free alternatives to traditional surgical skill development, offering precision, adaptability, and real-time feedback.
LSTM Model Evaluation
IBOM Lab, DGIST
