Innovation Forums 3

Industry Challenges of Efficient AI

Artificial intelligence (AI) is one of the core drivers of industrial development and a critical factor in promoting the integration of emerging technologies, such as graphic processing units (GPUs), Internet of Things (IOT), cloud computing, and the blockchain, in the new generation of big data and Industry 4.0. Efficient AI refers to the optimization of AI systems to perform tasks with minimal resource consumption, maximizing productivity while reducing costs and environmental impact. This encompasses advancements in algorithms, hardware, software, and data management techniques that ensure AI systems are not only powerful but also sustainable and accessible. The drive towards efficiency in AI is crucial in realizing AI’sits full potential, making intelligent solutions more widespread and integrated into everyday applications.

Moderator: Dr. Xin Chen

Machine Leaning Software Engineer, Intel

Bio:Dr. Xin Chen is currently a Machine Learning Software Engineer at Intel Corp. . Prior to joining Intel, he worked for Kuaishou US R&D Center, Media Emerging Technology Center (San Jose,CA), Mass General Hospital/ Harvard Medical School. His main work areas are computer vision, machine learning, hardware-software codesign for model compression, and high performance computing. Dr. Chen published top conferences such as CVPR, ECCV, IJCAI, ACM ICS,and ACL, and his work has been recognized with several awards, including Intel Division Achievement Awards and high rankings in international challenges.

 Dr. Chen received his Ph. D in Mechanical Engineering from University of Hawaii at Manoa in 2007. 

Speaker:  Sijia Liu

Professor of MSU

Bio: Sijia Liu is currently an Assistant Professor at the CSE department of Michigan State University, an Affiliated Professor at IBM Research, and an MIT-IBM Watson AI Lab affiliated PI. He received the Ph.D. degree (with All-University Doctoral Prize) in Electrical and Computer Engineering from Syracuse University, NY, USA, in 2016. He was a Postdoctoral Research Fellow at the University of Michigan, Ann Arbor, in 2016-2017, and a Research Staff Member at the MIT-IBM Watson AI Lab in 2018-2020. His research interests include scalable and trustworthy AI, e.g. scalable optimization for deep models, machine unlearning for vision and language models, AI robustness, and data-model efficiency . He received the Best Student Paper Award at the 42nd IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP’16), and the Best Paper Runner-Up Award at the 38th Conference on Uncertainty in Artificial Intelligence (UAI’22). He has published over 200 papers at top-tier ML/CV conferences, such as NeurIPS, ICML, ICLR, CVPR, ICCV, ECCV, AISTATS, and AAAI.

He is currently a Senior Member of IEEE, a Technical Committee (TC) Member of Machine Learning for Signal Processing (MLSP) in the IEEE’s Signal Processing Society, and an Associate Editor of IEEE Transactions on Signal Processing and IEEE Transactions on Aerospace and Electronic Systems, respectively. He has organized a series of Trustworthy AI workshops in ICML’22-’23 and KDD’19-’22, and provided tutorials on Trustworthy and Scalable ML in AAAI(’23, ’24), CVPR (’20, ’23, ’24), ICASSP’24, MLSP’23, NeurIPS’22, and KDD’19.

Speaker:  Jianhui Li

Sr. Principal AI Engineer, Intel

Bio: Jianhui Li is a senior principal AI engineer of Software and Advanced Technology group at Intel, leading deep learning compiler development. He has extensive experience in developing compilers, binary translators, and optimizing real-life applications for multiple domains. He started his career as assistant professor at Fudan university in 1997 and joined Intel in 2000 as software developer for binary translator, JIT compiler, XML processing performance libraries. One of his highlights is to initiate the Houdini product and lead the binary translator development to enable transparent execution of Android* ARM applications on Intel architecture. He invented oneDNN Graph API and contributed to open-source deep learning frameworks like Pytorch CPU module and Tensorflow PluggableDevice architecture. Graduated from Fudan University with a Ph.D. in Computer Science, he holds more than twenty U.S. patents and has published twenty academic papers.

Speaker:  Aosen Wang

CTO of Fengyun Vision Inc.

Bio: Aosen Wang, Ph.D. in Computer Science from the State University of New York at Buffalo, Bachelor from the University of Science and Technology of China. He is currently focused on applying large language models in vertical industries, including education, finance, e-commerce, and semiconductors. He has developed over ten related products, some of which have completed early market validation.

Previously, he worked at Apple's Silicon Valley headquarters in the FaceID core algorithm team, where he was responsible for developing the next-generation FaceID product for iPhone based on deep learning. After that, he worked at Meta (formerly Facebook) in the core machine learning algorithm department as a research scientist, focusing on developing the next-generation ad recommendation system.

With extensive experience in both academia and industry, Aosen has expertise in large language models, recommendation systems, deep learning, computer vision, and smart health. He has published over 30 papers in academic conferences and top journals, has won multiple best paper awards and nominations, and holds several U.S. patents in recommendation system architecture and algorithms. He has conducted in-depth research on the application of large models, recommendation system optimization, and user profile understanding. Additionally, he has achieved recognition in the miniaturization and localization of general neural network models, winning awards in international competitions such as CVPR and obtaining U.S. patent for neural network quantization. His other research on using artificial intelligence to improve health, including disease detection and monitoring, has garnered strong industry response and collaboration interest.

Speaker:  Hanxian Huang

Ph. D Candidate of UCSD

Bio: Hanxian Huang is a fifth-year PhD Candidacy in CSE at the University of California San Diego, advised by Prof. Jishen Zhao. Before joining UCSD, she received her B.S. in EECS at Peking University. Her research interests broadly span the intersection of machine learning with programming languages, compilers, and computer systems, as well as co-designing efficient machine learning algorithms and systems. Hanxian was selected as one of the Machine Learning and Systems Rising Stars 2024 by ML Commons and she will be in the job market in academia and industry in the fall of 2024.

Speaker:  Mostafa El-Khamy

Sr. Principal Engineer of Samsung  Electronics America

Bio: Mostafa El-Khamy (S'01---M'07---SM'12) received his Ph.D. and M.S. degrees from the California Institute of Technology (Caltech), USA, and his M.S. and B.S. degrees from Alexandria University, Egypt, all in Electrical Engineering. He received his MBA from the Edinburgh Business School, UK. He is a Senior Principal Engineer with Samsung DSRA (Device Solutions Research America).  He is also the Co-Chair of the MLPerf Mobile Working Group for AI benchmarking. Dr. El-Khamy is an Adjunct Professor at the Faculty of Engineering, Alexandria University, and was a founding faculty member of Egypt-Japan University for Science and Technology (E-JUST). Previously, he was at Qualcomm R&D, San Diego. His research interests include the theory and practice of artificial intelligence in multimedia and communication systems. He is the recipient of the URSI Young Scientist Award, the Caltech Atwood Fellowship, the Alexandria University Scientific Incentive Award, the Samsung Best Paper Award, and the Samsung Distinguished Inventor Award.