*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
Title: Energy-efficient Image Processing for Intelligent Sensor Systems
Committee:
Dr. Mukhopadhyay, ECE, Advisor
Dr. Raychowdhury, ECE, Chair
Dr. Yalamanchili, ECE
Abstract:
The objective of the proposed research is to design energy-efficient image processing algorithms and architecture for intelligent wireless image sensor systems. For reliable delivery of region-of-interest (ROI) under dynamic environment, the research explores low-power moving object detection with enhanced noise robustness. Based on the ROI information, the system energy is further optimized by a low-power ROI-based coding scheme and an on-line data rate controller. To enable machine learning based intelligent image processing with limited hardware resources, the research proposes neural network design with lower storage and computation demand. The storage demand is reduced by compressing the neural network weights with an adaptive image encoding algorithm. The computation demand of convolutional neural networks is optimized by mapping the entire network parameters and operations into the frequency domain. FPGA verification and analysis of on-chip neural network inference will complete this research, enabling intelligent image processing on resource-constrained mobile sensor platforms.
