1. Professor NGO Chong Wah
(Singapore Management University (SMU))
Dr. Chong-Wah Ngo is a Professor in the School of Computing and Information Systems at Singapore Management University (SMU). He received his PhD in Computer Science from the Hong Kong University of Science and Technology. Prior to joining SMU, he was a Professor at City University of Hong Kong. His research interests include multimedia retrieval, multimodal learning and fusion, computer vision, and video understanding. Dr. Ngo currently serves as an Associate Editor of ACM Transactions on Multimedia Computing, Communications, and Applications (TOMM). He is the General Co-chair of the International Conference on Multimedia Retrieval (ICMR 2027) and previously served as General Chair of The Web Conference (WWW 2024). He has also served as Program Co-chair of ACM Multimedia Asia 2025 and ACM Multimedia 2019.
Talk: What Multimodal Large Language Models See and Miss in Cultural Contexts?
Multimodal large language models (MLLMs) have made substantial progress in image and video question answering (QA). Yet their ability to understand cultural content beyond what is directly visible remains largely underexplored. Cultural understanding requires reasoning over implicit knowledge, traditions, social norms, and temporal context, rather than simply recognizing objects, scenes, or actions. The talk will examine the capabilities of MLLMs in grounding visual evidence for image and video cultural QA, highlighting both their strengths and current limitations. The talk will discuss the challenges in culture-aware visual understanding and outline promising research directions toward context-aware multimodal reasoning. Two new benchmarks, Seeing Culture Benchmark (SCB) and the Cultural Moment Benchmark (CMB), designed to evaluate the reasoning capabilities of MLLMs on Southeast Asia (SEA) cultures, will also be introduced.
