Musical Key Recognition using a Hidden Markov Model
Every musical piece has a distinct key. In this project, a Hidden Markov Model is trained to try and predict such a key.
Backdoors in Neural Networks
Demonstrating a 'backdoor' security vulnerability. Trained networks in PyTorch and MXNet with ImageNet dataset on a supercomputer cluster. Demo with React.js app, using ONNX.js on-device inference. ✩ Project was rated 'best of class 2020'
A fusion between a deep-learning and a filter-based approach for road marker detection
Fusing the outputs of a supervised deep-learning and an unsupervised filter-based road marker detector into a new supervised model, reaching 95% accuracy.
