May 15, 2014
EECS Doctoral Dissertations

Energy Efficient System Design

Rahul Rithe, MIT
  • Speaker
  • Abstract
  • Exclusive Content

Speaker PhotoRahul Rithe received the B.Tech. (Honors) degree in electronics and electrical communication engineering from the Indian Institute of Technology, Kharagpur, India, in 2008, and the S.M. and Ph.D. degrees in electrical engineering and computer science from the Massachusetts Institute of Technology, Cambridge, USA, in 2010 and 2014 respectively. He received the best B.Tech. thesis award at IIT Kharagpur in 2008 and the best S.M. thesis award in electrical engineering at MIT in 2010. His research interests include low-power integrated circuits and energy-efficient algorithms and architectures for portable multimedia applications. Dr. Rithe was the recipient of the President of India Gold Medal and the MIT Presidential Fellowship in 2008.

Advances in portable computing through applications such as enhanced video, imaging and graphics performance and the ability to perform continuous non-invasive health monitoring are essential to making portable devices platforms for productivity and healthcare as well as entertainment. The high computational complexity of such applications necessitates system level innovations - including algorithms, architectures and circuit design - to enable “Moore’s Law” for energy-efficiency. This presentation discusses two portable computing applications - a reconfigurable processor for computational photography and a portable platform for medical imaging.

Computational photography applications have so far been software based, which adds significant complexity and does not support real-time processing. The reconfigurable processor implements High Dynamic Range imaging, Low-Light Enhancement and Glare Reduction. Power reduction techniques at all stages of the design, including algorithmic optimizations, highly parallel architecture and circuit design for low-voltage operation, significantly enhance energy-efficiency compared to software implementations on recent mobile processors, while enabling real-time processing.

Advances in computational photography and computer vision, coupled with efficient processing on portable multimedia devices, provide a unique opportunity for portable medical imaging. Extending and enhancing computer vision techniques, such as level set method based image segmentation, feature detection and registration, enable a portable platform for skin lesion detection and progression analysis of skin conditions. Algorithmic optimizations lead to significant performance enhancements and pave the way for energy-efficient hardware implementations for medical imaging.

 

This content is restricted to our MIG members and members of the MIT community. Login below, or contact us for more information about our partner programs.

Login