Vasu is presently pursuing his Masters in Language Technologies from the School of Computer Science at CMU. His primary areas of interest are Multimodal Machine Learning, Dialog Systems, Question Answering and building explainable and robust Deep Learning models via adversarial training.
He is presently working on trying to decode and reconstruct the Neural Basis of Real World Social Perception by trying to find relations between intra-cranial EEG signals and human expressions, pose, actions, emotions etc. He is also working on designing adversarial attack to test the robustness of Visual Question Answering Models and build training pipelines which make these models robust to such adversarial attacks. He is also interested in various facets of facial expression analysis.