Project Abstract

I decided to do my project on image classification using CNN’s (Convolutional Neural Networks) as I believe it to be an interesting and challenging topic. As for why finding Wally, I believe it to be a well suited task to test my ability and understanding of the topic as well as inform me on the complexity of performing certain tasks when applied to other fields. When starting this topic my main aim was to see how plausible it is to make a working image classifier that can distinguish wally out of the background to a reasonable degree of accuracy. My main findings to this point is that is with the smaller amount of data available to train the CNN it is hard to get a high degree of accuracy when training the model, however I hope to find that with methods such as augmenting the data and data regularization techniques i will still be able to get the model performing well.

Keywords: Image Classification, Artificial Intelligence, Convolutional Neural Networks

 Conference Details

Session: A

Location: Sir Stanley Clarke Auditorium at 11:00 13:00

Markers: Cécilia Pradic, Jay Morgan

Course: BSc Computer Science 3yr FT

Future Plans: I’m looking for work