COLLEGE OF COMPUTER SCIENCE NCTU
Topic: Smartening edge devices with neural network accelerator ASICs
Speaker: Youn-Long Lin
Chair Professor of Computer Science
National Tsing Hua University, Taiwan
Time: 5/8 (Wed) 3:30-5:00PM
Place: ED 117 (工四117)
Recent rapid advancements of deep learning artificial neural network technologies and applications have raised an urgent issue of augmenting IoTs (Internet of Things) with cognitive capabilities. However, NN models developed/trained in the cloud/server cannot be deployed into edge devices due to size/power/bandwidth constraints. In this talk, we describe how to take hardware implementation into account during NN model development/training and how to take advantage of massive parallel processing using advanced semiconductor processing technology. We will illustrate the hypothesis using a case of video semantic segmentation.
Youn-Long Lin is a professor with Department of Computer Science, National Tsing Hua University. He is a co-founder of Global Unichip Corp. He received his B.S. in electronics engineering from National Taiwan University of Science and Technology in 1982 and Ph.D. in computer science from the University of Illinois, Urbana-Champaign, in 1987. His research areas include electronic design automation, VLSI design for video coding and neural network acceleration.