Building models that learn spoken language by seeing and hearing
We aim to learn language by distant supervision through captioned videos, similarly to how children learn language through interacting with the world around.
Our goal is to develop new tools for modeling diverse multi-agent settings, and design estimation algorithms to unravel the strategic interactions among the agents.
Our goal is to design novel data compression techniques to accelerate popular machine learning algorithms in Big Data and streaming settings.
We develop algorithms for fetal MRI interpretation, to enable noninvasive fetal monitoring from MRI.
AIRvatar is a system that telemetrically collects and analyzes fine-grained data on users’ virtual identities and the process used to create them.
From audio recordings of clinician-subject interactions, we determine the spoken language bio-markers that are associated with health outcomes, such as dementia and depression.
Traffic is not just a nuisance for drivers: It’s also a public health hazard and bad news for the economy.
A "precision medicine" approach for finding relevant cancer treatments in clinical literature and eligible trials. For a given patient with associated demographics (age, gender) and disease (cancer type, genetic variants), we query a database of all pubmed articles and clinicaltrials.gov trials using NLP techniques to find the most useful and relevant treatments for the patient. Our ensemble-based system performed very well in the TREC 2016 Precision Medicine challenge.
We are designing new parallel algorithms, optimizations, and frameworks for clustering large-scale graph and geometric data.