| Prof. Hai LiuCentral China Normal University, China Hai Liu is a Professor and PhD Supervisor at the Faculty of Artificial Intelligence in Education, Central China Normal University. He received his Ph.D. degree in Pattern Recognition and Intelligent Systems from Huazhong University of Science and Technology. His long-term research interests cover self-regulated learning, learning resource recommendation, knowledge graphs, computer vision, and machine learning. Dr. Liu was selected for the "Hong Kong Scholar" talent program and conducted a two-year research visit at the Robotics and Vision Laboratory, City University of Hong Kong. He has presided over 6 projects funded by the National Natural Science Foundation of China (NSFC) and participated in 8 others. He has published more than 120 academic papers in prestigious domestic and international journals and conferences, among which over 60 SCI/SSCI/CSSCI journal papers were published as the first or corresponding author, including 25 IEEE Transactions papers in CAS Zone 1. Eighteen of his papers have been selected as ESI Highly Cited Papers and 12 as ESI Hot Papers. He has filed more than 80 national invention patents, with over 40 granted. Dr. Liu was included in the World's Top 2% Scientists 2025 list released by Stanford University. Since 2022, he has served as a review expert for the Young Changjiang Scholars Program of the Ministry of Education. He has been awarded the First Prize of Hubei Science and Technology Progress Award (2020) and the First Prize of Science and Technology Progress Award (Scientific and Technological) in the Scientific Research Outstanding Achievement Award for Institutions of Higher Education (2019).。 Speech Title: Research on Key Technologies for Fine Estimation of Learners' Head Postures in Intelligent Learning Evaluation Abstract: In this talk, we present our research on key technologies for fine-grained head pose estimation of learners for intelligent learning evaluation. Accurate head pose estimation serves as a core upstream technique for perceiving learners’ attention status and evaluating classroom engagement in intelligent learning evaluation scenarios, which is of great research value and practical significance for improving the precision of teaching evaluation and the level of personalized instruction. However, complex challenges prevalent in real classroom scenarios, including facial occlusion, large-angle head deflection and uneven illumination, cause deformation and distortion of facial features, severely restricting the accuracy and robustness of pose estimation and becoming a key bottleneck for the deployment of intelligent learning evaluation. Over the past decade, our team has carried out systematic research on head pose estimation. Built upon the state-of-the-art Transformer architecture and integrated with technologies related to large model representation and educational scenario adaptation, we thoroughly characterize the morphological deformation laws of different facial regions during head deflection and construct a morphology-aware relational modeling framework, effectively overcoming core difficulties such as occlusion and large-angle pose estimation. Our methods achieve leading performance on three authoritative international benchmark datasets: BIWI, AFLW2000 and 300W-LP, and the series of research results have been published in more than ten papers in top international journals and conferences including IEEE TIP, CVPR, IEEE TII, IEEE TMM and IEEE TCSVT, providing reliable technical support for learner pose perception in intelligent learning evaluation scenarios. |
| Prof. Songhua DuGuangdong University of Technology, China Du, Helen S. graduated from City University of Hong Kong in 2008 with a Ph.D. in Information Systems. Professor of the Department of Data Science and Business Intelligence, School of Management, and Doctoral Supervisor. Also serves as the Director of the International Cooperation and Accreditation Center of the School of Management. Special Appointed Professor of the "Hundred Talents Program" of Guangdong University of Technology (2014-2018). With extensive teaching and research experience in the fields of e-commerce and information management. Currently, the main research topics include behavioral analysis empowered by digital intelligence, human-computer and human-intelligence interaction, gamification design, pro-environmental behavior, and green innovation and entrepreneurship models. Relevant research achievements have been selected as typical case study results by the Ministry of Education in 2023 and have been highly cited in ESI. Has led 3 national research projects, 3 provincial research projects, and 1 project of the Guangdong Province Graduate Education Innovation Program, etc. Has published over 40 academic papers in authoritative domestic and international journals such as International Journal of Information Management, Decision Support Systems, Journal of Cleaner Production, Internet Research, Journal of the Association for Information Science and Technology, International Journal of Human-Computer Studies, Nankai Business Review, and Operations Research and Management. Currently serves as a member of the editorial board of the SSCI journal Online Information Review, associate editor of the EI journal Journal of Information & Knowledge Management, and holds social positions such as a member of the Economic Committee of the Guangdong Provincial Committee of the China Democratic League and a member of the Teaching Steering Committee for E-commerce Majors in Undergraduate Colleges of Guangdong Province. In addition, I have many years of international practical experience in enterprise information management. I have worked as an information system analyst for several years at Ontario Power Generation in Ontario, Canada and PCCW in Hong Kong. |
