My First Publication

I'm extremely proud to say that this is my first published paper. My role in this research project was to evaluate the relationships within scene graphs and test different data pruning methods. To evaluate these methods, model performance in both training speed and accuracy on unseen images were tested and recorded. If you would like to read the paper for the full details and scope of this research, please visit the publication here.

Key Achievements:

1. Q1 Publication: Getting a Q1 publication in a respected journal is a dream for any academic, let alone one that is still in University.
2. Developing A Novel Approach To VQA: During the course of this research, we developed a novel technique when pruning known scene graph data to improve overall model accuracy, and as a byproduct, training times.

Head over to my github to see the codebase