Dr Xiao Hua Wu
Ms Yee Wai Yim
Dr Cheung Ming Lai
Mr Ngai Yat Cheung
Mr Chi Chung Leung
Ms Yin Yee Chan
Mr Chun Chung Lo
Mr Sheng Peng Wu
Mr Fu Chun Li
Though mobile devices are more becoming the main factor of computing, most of the smart device’s computational and storage capability are limited. In order to achieve better Quality of Experience (QoE) particularly for the computational intensive multimedia processing applications such as content retrieval, object recognition, and virtual/augment reality, intelligent integration between cloud and mobile computing becomes critical. Mobile edge computing (MEC) is the emerging technology for IoT and 5G, where the edge cloud could provide better QoE for low-latency, location aware applications. In this project we will focus on the feasibility study of algorithm partition and dynamic computational offloading for mobile multimedia processing, the major algorithm includes feature extraction, image matching, smoothing and anti-aliasing. We will partition and optimize algorithm modules on CPU and GPU for both mobile device and edge cloud, make offloading decision to adapt dynamic changes in network, edge device, resource and mobile devices to achieve enhanced QoE. Compute intensive computer vision algorithm partition SDK and compute offloading middleware will be designed to support selected mobile multimedia applications.