Kevin works on multi model learning using deep learning and convolutional neural networks. He is also working on expanding the capabilities of Yale Image Finder by applying deep learning methods to classify images from PubMed publications.