Lei Yang Wins 2021 IEEE TCAD Best Paper Award
ECE Asst. Prof. Lei Yang has been selected by the IEEE Transactions on Computer-Aided Design (TCAD) for the 2021 Donald O. Pederson Best Paper Award, presented at the Design Automation Conference in December 2021.
The TCAD award recognizes the best paper published in the Transactions on Computer-Aided Design of Integrated Circuits and Systems publication.
The IEEE TCAD Editorial Board and the IEEE Council on Electronic Design Automation selected the paper as one of two 2021 winners from over 800 papers published by the journal in the last two years.
The winners were judged on the overall quality, originality, level of contribution, subject matter, and timeliness of the research. The award was presented at the Design Automation Conference in December 2021.
Dr. Yang's paper, “Hardware/Software Co-Exploration of Neural Architectures,” proposes a novel hardware and software framework for efficient neural architecture search (NAS), a technique to automate machine learning systems.
"This is the first work for the co-exploration of hardware design space and neural network architecture search space," said Dr. Yang.
"Compared with the existing Hardware-aware NAS, in this work, our proposed co-exploration framework has demonstrated the best tradeoff between the performance of neural network architectures and the requires from hardware platforms."
Dr. Yang joined UNM after her experience as a Post-Doctoral Research Associate in the Department of Computer Science and Engineering at the University of Notre Dame. Before that, she was a research scholar at the University of California, Irvine from Oct. 2017 to Feb. 2019, and a research scholar at the University of Pittsburgh from Feb. 2019 to Aug. 2019.
Dr. Yang's research focuses on Automated Machine Learning, Embedded Systems, and High-Performance Computing Architectures. Dr. Yang has created methods in the Hardware/Software co-exploration for neural architectures. Approaches developed by her group have been deployed to the resource-constrained edge AI systems and applied to medical applications to improve the quality of healthcare and largely reduce cost. Currently, Dr. Yang is also working on System-Level Optimization for Applied Machine Learning, and designing optimized machine learning models for specific embedded systems and applications, including fair medical AI, collaborative drug discovery, and subsurface estimation.
Comparison between (a) hardware-aware NAS; (b) the proposed hardware/software co-exploration NAS. The red rectangles convey the metrics that can be optimized in the exploration.