Zeeshan Ashraf is a graduate student at Carnegie Mellon University. His research interests lie in multimodal machine learning, computer vision and generative modelling. Currently he is also working on low resource machine translation and VQA using unstructured knowledge. He looks to learn more about and undertake research in Bayesian Learning. Previously, he has worked on gradient estimation in Stochastic Neural Networks at International Institute of Information Technology, Hyderabad.