Final Year Project: Machine Learning and Simulation of GPR Data
Final year project where I worked with Ground Penetrating Radars (GPRs) to collect and synthesise tree root scans using Generative Adversarial Networks (GANs). By augmenting the GPR scans, this project aims to address the lack of data used for tree health monitoring and insights. Through this project, I experienced the full journey of machine learning, from data collection and preprocessing to model training and reiteration. By implementing and comparing different GAN architectures (DCGAN, WGAN-GP, pix2pix), I learned the nuances of making these models work, from fine tuning parameters and learning rates to recognising signs of mode collapse.
Machine Learning
Generative Adversarial Networks
Data Collection
Data Preprocessing
Data Augmentation
Python
PyTorch
TensorFlow
MATLAB
Research and Analysis