This talk presents holistic approaches to realize energy-optimized, machine-learning-enabled Internet-of-Things (ML-IoT) systems and VLSI circuits. The optimized system integration is a major challenge in intelligent IoT systems. A truly energy-optimal ML-IoT solution is attainable only by a cross-layer optimization that requires a full characterization of the complete end-to-end system. Addressing this critical technical challenge in emerging ML-IoT applications, a cross-layer interdisciplinary research that spans deep learning algorithms, wireless communication, digital signal processing, and VLSI hardware architecture is discussed in this talk. Presented research projects aim for ultra-low power and/or energy-aware wireless IoT systems enabled by novel hardware-friendly algorithms and cross-layer optimized VLSI systems.
Hun-Seok Kim is an assistant professor at the University of Michigan, Ann Arbor. Kim received his B.S. degree from the Seoul National University (South Korea) in 2001, and M.S. & Ph.D. degrees from the University of California, Los Angeles (UCLA), all in Electrical Engineering. His research focuses on system analysis, novel algorithms, and efficient VLSI architectures for low-power/high-performance wireless communication, signal processing, computer vision, and machine learning systems. Before joining the University of Michigan, Kim worked as a technical staff member at Texas Instruments (2010 – 2014). He is serving as an associate editor of IEEE Solid State Circuits Letters, IEEE Transactions on Mobile Computing, and IEEE Transactions on Green Communications & Networking. Kim is a recipient of the 2018 Defense Advanced Research Projects Agency (DARPA) Young Faculty Award (YFA) and the National Science Foundation (NSF) Faculty Early Career Development (CAREER) Award 2019.