brief bio

As of Nov. 2011, I joined the Interactive Visual Media group at MSR Redmond. Before joining MSR I helped co-found Anchovi Labs (acquired by Dropbox in 2012) and was a postdoc at the Computation Vision Lab at Caltech under Pietro Perona. I received my PhD under the guidance of Serge Belongie at UCSD in 2007 and have continued doing research in computer vision and machine learning since.

My research interests lie in machine learning and pattern recognition and their application to computer vision. In particular, I aim to answer two fundamental questions: What are the most effective representations of images and videos that make subsequent reasoning about their content tractable? And what learning paradigms optimally capture the structure and properties of visual information? Understanding the interplay between representation and learning is central toward achieving a more complete understanding of machine vision.