报告人：陈仲澍 (Chung Shue Chen)

报告地点：计A505

(一）

报告题目：The Deep Learning Revolution - An introduction

报告时间：8月20日 下午13:30-14:30

报告摘要：Machine learning has become one of today’s most important and rapidly growing fields of technology, with applications increasingly appearing across a wide range of scientific and engineering domains. As a powerful branch of machine learning, deep learning provides a general-purpose framework for learning complex patterns and representations directly from data through computational models known as neural networks. Recent advances in deep learning have enabled remarkable progress in many challenging problems and have also led to the emergence of powerful generative artificial intelligence systems and large language models. In this talk, we will introduce the fundamental concepts and terminology of machine learning and deep learning and illustrate their broad applicability through several representative examples. In particular, we will discuss applications of deep learning to medical diagnosis, where neural networks can learn to classify skin lesions from images; protein structure prediction, where deep learning models predict three-dimensional protein structures from amino acid sequences; image synthesis using generative models; and large language models for natural language processing and generation. We will also briefly review the development of neural networks and introduce the fundamental framework underlying modern deep learning, including feed-forward neural networks, gradient-based optimization, and error backpropagation. Finally, we will discuss how advances in model scale, training data, computational resources, and neural network architectures have contributed to the deep learning revolution and the remarkable capabilities of today’s deep learning systems.

（二）

报告题目：AI-assisted Integrated Semantic Sensing and Communication for NTN-enabled 6G Networks–An introduction

报告时间：8月20日 下午15:00-16:00

报告摘要：Non-terrestrial networks (NTNs) are expected to play an important role in future 6G networks by extending connectivity beyond the coverage limitations of conventional terrestrial networks. Recent advances in satellite technologies, particularly Low Earth Orbit (LEO) satellite communications, are accelerating the convergence of terrestrial and non-terrestrial networks and enabling ubiquitous and resilient connectivity for mobile and IoT devices. Beyond conventional communications, integrated terrestrial and non-terrestrial networks also provide new opportunities for positioning, object tracking, localization, and sensing, which are essential for emerging applications such as autonomous vehicles, smart cities, industrial Internet of Things, emergency response systems, and augmented reality. In this talk, we will first introduce recent developments in satellite communications and NTNs, as well as their evolution toward integrated terrestrial and non-terrestrial 6G networks. We will then discuss the use of radio signals for object tracking, localization, and sensing, with particular attention to three-dimensional position, velocity, and orientation estimation. Furthermore, we will introduce an AI-assisted integrated semantic sensing and communication framework for NTN-enabled 6G networks. To address the limited bandwidth, high propagation latency, and dynamic link conditions of NTNs, the proposed framework exploits semantic communication to transmit high-value semantic information rather than raw data, thereby reducing bandwidth requirements, accelerating decision-making, and improving service quality. Finally, we will discuss how Vision-Language Models (VLMs) can further enhance semantic sensing by exploiting rich vision-language representations and supporting a wide variety of visual understanding and decision-making tasks, highlighting their potential for intelligent sensing and communication in future 6G networks.

报告人简介：

Chung-Shue (Calvin) Chen is a Senior Research Scientist and DMTS at Nokia Bell Labs. He also holds a position of Permanent Member at LINCS. Prior to joining Bell Labs, he worked at INRIA, in the research group on Network Theory & Communications (TREC, INRIA-ENS). He was an ERCIM Alain Bensoussan Fellow at the Norwegian University of Science & Technology (NTNU, Department of Electronics & Telecommunications) and the National Center for Mathematics & Computer Science (CWI, Amsterdam). He was an Assistant Professor at The Chinese University of Hong Kong (CUHK). He worked at CNRS in Lorraine on Real-Time & Embedded Systems. He has served as TPC in international conferences including ICC, Globecom, WCNC, INFOCOM, PIMRC, VTC, CCNC and WiOpt (TPC Vice Chair). He is an Editor of European Trans. on Telecommunications (ETT) and an Associate Editor of Telecommunication Systems (Springer). He is an IEEE Senior Member. His research spans a wide range of topics in information & communication engineering, 5G/6G, IoT, wireless network, indoor positioning, resource allocation & scheduling problem, intelligent system, and learning algorithm. He was a co-recipient of the Best Paper Award from ACM MobiCom S3 Workshop, the Excellent Paper Award from ICUFN, and the Optoelectronic Technology Innovation Award from IEEE OGC.