Research

Recognising specific foods in MRI scans using CNN and visualisation

Published to Eurographics CGVC 2020. Research into recognising specific foods in MRI scans using CNN

Detecting specific foods in MRI scans using TensorFlow and RetinaNet implemented in Python

My undergraduate dissertation in researching object recognition and training a CNN with RetinaNet to detect specific food in MRI scans of the stomach

Projects

eigenFaces facial recognition

Written in: MATLAB
Facial recognition using PCA/eigenfaces for dimensionality reduction to recognise if a face is in a dataset. As well code to reconstruct images, and compare variance for better understanding of PCA/eigenfaces

ImageApp

Written in: C++, QT 5 GUI, OpenCV 4.4.
Intended to expand my knowledge of C++ with UI, incorporating my skills in OpenCV to create an image editing application much like paint.net & GIMP. Current Features: Read/Write images, basic editing and image effects.
Currently under hiatus due to academic studies.

Optimisation of machine learning algorithms

Written in: Python, juypter notebook, Scikit-Learn.
A part of my ML course for my postgrad where I had to produce a report on testing & comparing various classification algorithms (Random Forests, SVM, Naive Bayes) on the fashion MNIST dataset and optimise the hyperparamters to get the best accuracy.

Gwyendd Closest Postcode

Stack: Python, SQLite
Part of a group project where I developed the backend functionaility and database interactions. Intended to be client work to find the nearest GP's, Schools, dentists etc by inputting the users postcode, GeoPy was then used to calculate the nearest GP. SQLite was used for the DBMS as the brief didnt require a server-client setup so using an on disk database was the best solution.

Detecting specific foods in MRI scans using TensorFlow

Stack: Python, TensorFlow
The work I did for my dissertation using a open source implementation of Facebooks RetinaNet CNN. The work was part of a multidisciplinary research project into the study of digestion. The work presented used Deep Learning object recognition technics to detect peas in stomach MRI.
Other skills that were needed were automated file manipulation, XML, CSV parsing and image editing using pythons built in libaries like pandas, matplotlib, numpy, opencv.

Java Projects

Written in: Java 8, Maven, Junit, Swing, Servlets
An assortment of java projects I made for my undergrad modules, including: Database CLI, Chat Client, weather forecast app, pong game.

Python Projects

Written in: Python 3
An assortment of python projects of my own work, examples include: Movie organiser, integer to roman numeral and matrix covariance calculator