March 3, 2023 Friday
Abstract:
Over the past few years, Deep Neural Networks (DNNs) have enabled remarkable breakthroughs in a variety of scenarios, including autonomous driving, extended reality (XR), and view synthesis. Mobile devices, with their power-efficient and specialized processors and real-time scenario suitability, are becoming the primary carriers for these emerging applications. Recently, as a result of AutoML tools (e.g., Network Architecture Search) and other training-related advancements, DNNs are designed to be deeper with increasingly complex structures and larger computation sizes. Given the increasing computational demands, real-time DNN execution is an ideal but difficult goal for mobile devices due to the limited computing and storage resources.
In this talk, I will present our recent efforts to achieve real-time DNN inference on mobile devices with compiler optimizations. It mainly focuses on three innovative aspects: (1) a compression-compilation co-design framework for computation-intensive models; (2) an advanced operator fusion framework for extremely deep (memory-intensive) models; and (3) our latest work on global data layouts and low-level instruction optimizations for emerging power-efficient processors (mobile DSP). In the end, I will describe my vision for the future of real-time machine learning systems on heterogeneous platforms as well as my future research directions.
Speaker Bio:
Mr. Wei NIU
PhD Candidate, Department of Computer Science, Collegeof William & Mary
Wei Niu is a fifth-year Ph.D. candidate in the Department of Computer Science at William & Mary. Wei’s research interests lie in real-time machine learning systems, mobile computing, parallel computing, and compilers. In particular, he focuses on achieving real-time DNN execution on mobile platforms with compiler optimizations. His work has appeared at top conferences (e.g., MICRO, PLDI, ASPLOS, RTAS, ICS, DAC, NeurIPS, CVPR, AAAI, ECCV, ICCV) and top journals (e.g., TPAMI, CACM). He is the recipient of the Stephen K. Park Graduate Research Award at William & Mary. He also won first place in the 2020 ISLPED Design Contest, the CACM Contributed Article Award in 2021, and the Best Paper Award at an ICLR workshop in 2021. Previously, he earned his bachelor’s degree from Beihang University in 2016.