Dr. Michele Covell

Dr. Michele Covell studied at University of Michigan (BSEE) and MIT (MS, PhD).  Since then she has been working in the areas of audio and video processing and machine learning at various research labs, including SRI International, Interval Research Corporation, HP Labs, and Google Research.  In 2006, HP Labs included her team's work on streaming content-distribution networks for mobile as one of the "top 40 accomplishments in HP Labs' [then] 40-year history".  After joining Google, Michele worked on audio and video fingerprinting, technology that is still used for copyright protection by YouTube.  She received awards for this work for technical innovation and financial impact and the YT "Content Id" System received a Technical Emmy in 2013.  In her later work, she has explored ways to combine machine learning with data compression, from neural networks for H.264 parameter control; to very low bit-rate image compression using stochastic gradient descent for encoding and neural networks for decompression; to end-to-end network based image compression.  Her latest efforts have been related to music processing.