From Compression to Transformation: Efficient Deep Learning on Mobile Devices and LLMs in Education

Abstract

Driven by advancements in artificial intelligence (AI), an increasing number of intelligent applications are emerging on mobile devices. As one of the most representative AI technologies, Deep learning has been successfully applied in various domains such as computer vision, speech recognition, natural language processing, and audio recognition, achieving performance on par with, and sometimes surpassing, that of human experts. However, the high computational, memory, and energy requirements of deep learning models present significant challenges for deployment on resource-limited mobile devices. This seminar will explore cutting-edge research on efficient deep learning algorithms for on-device machine learning through deep learning compression. We will address the challenges of deploying deep learning models on mobile devices, and investigate strategies to optimize computational efficiency, memory usage, and energy consumption, while still achieving high-performance results. In addition, the seminar will delve into the transformative role of Large Language Models (LLMs) in education, highlighting their potential to reshape content generation, personalized tutoring, and student engagement. By integrating LLMs with traditional educational models, the seminar will demonstrate how these technologies can overcome existing challenges in content creation, accessibility, and scalability, offering a forward-looking perspective on how AI and LLMs can drive both technological innovation and educational transformation.

Speaker Biography

Dr. Zhuwei Qin is an assistant professor in Computer Engineering and the Director of the Mobile and Intelligent Computing Laboratory (MIC Lab) at SFSU. He has broad research experience in efficient deep learning computing, and edge computing specifically, in researching computational optimization for accelerating deep learning on low-power mobile and edge devices. His research has been funded by multiple state-wide academic and industrial programs such as the CSU STEM-NET Faculty Interdisciplinary Collaborative Research SEED Grant Program and the Sony Sensing Solution University Collaboration Program (SSUP). He is also a co-PI on an NSF IUSE project focused on AI-driven support for neurodiverse teamwork. Dr. Qin teaches both undergraduate and graduate courses in AI, including “On-Device Machine Learning” and “AI in Engineering” at SFSU.