The automatic detection and classification of bugs in code blocks committed to repositories is invaluable in software development. Traditionally, this issue has been handled through semantic analysis. Recently, a new line of research has emerged suggesting that computer vision can be used for the same task; i.e. analyzing digital images and/or screenshots of code in order to determine the presence/type of bugs! (https://dl.acm.org/doi/pdf/10.1145/3485135)
In this project you are expected to work with existing datasets and implement a variety of deep convolutional neural network architectures for the detection and classification of software bugs.