Keynote

Keynote 1 (Sept. 10, 9:00-10:00)
Towards a Safer Future: Empowering Secure AI

Abstract

As AI systems, especially in computer vision, become increasingly integrated into our daily lives, ensuring their security is critical. This talk addresses key vulnerabilities and defense strategies for robust and trustworthy AI. We explore adversarial attacks, including our research on physical adversarial examples that can fool AI in real-world settings, and present advances in face anti-spoofing to counter adversarial attacks. We examine the growing risks of deepfakes, highlighting our recent work in image editing with large text-to-image models, and discuss both the creative potential and misuse concerns. The talk also covers hallucination in multimodal large language models (MLLMs), presenting solutions to improve factual accuracy and reliability. Finally, we look ahead to the emerging era of agentic AI, where autonomous, goal-driven agents introduce new security challenges. We outline design principles and mitigation strategies to build resilient AI that can withstand evolving threats. By combining technical insights, case studies, and forward-looking perspectives, this talk aims to inspire the development of secure AI systems that can be confidently deployed in real-world applications.

Biography

Wen-Huang Cheng is a University Distinguished Chair Professor in the Department of Computer Science and Information Engineering at National Taiwan University and a Visiting Professor at the Korea Advanced Institute of Science and Technology (KAIST). His current research interests include multimedia, computer vision, and machine learning. He has actively participated in international events and played significant leadership roles in prestigious journals, conferences, and professional organizations. These roles include serving as Editor-in-Chief for IEEE CTSoc News on Consumer Technology, Senior Editor for IEEE Consumer Electronics Magazine (CEM), Associate Editor for IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) and IEEE Transactions on Multimedia (TMM), General Chair for ACM MMAsia (2023), IEEE ICME (2022), and ACM ICMR (2021), Technical Program Chair for ACM MM (2025), ACM ICMR (2022), IEEE ICME (2020), IEEE VCIP (2018), Chair for IEEE CASS Multimedia Systems and Applications (MSA) technical committee, and governing board member for IAPR. He has received numerous research and service awards, including the NVIDIA Academic Grant Program Award (2025), the 2024 Best Paper Award of IEEE Consumer Electronics Magazine, the Best Paper Award at the 2021 IEEE ICME and the Outstanding Associate Editor Award of IEEE TMM (2021 and 2020, twice). He is an IEEE Fellow, IET Fellow, and ACM Distinguished Member.

Keynote 2 (Sept. 11, 9:00-10:00)
Transforming Animal Behavioral Research with AI and IoT Technologies

Abstract

In this talk, we present our recent efforts to advance the study of animal behavior through the development of AI- and IoT-assisted systems. Biologists commonly attach sensor loggers, such as GPS units, accelerometers, and cameras, to animals to record their movements and activities, generating vast behavioral datasets. However, two key challenges arise: designing advanced sensing devices that comply with strict weight constraints for animal-mounted deployment, and extracting meaningful insights from the resulting behavioral big data.

To address the first challenge, we introduce our AI-enabled bio-loggers, which autonomously detect and record valuable animal behaviors. By incorporating intelligent power management, these devices can function with smaller batteries, enabling a lightweight design that minimizes impact on the animals. To address the second challenge, we present a deep learning-assisted platform that supports comparative analysis of animal movement data. Leveraging attention-based neural networks, this system highlights trajectory segments that are characteristic of specific groups, helping researchers uncover key behavioral patterns and generate new hypotheses. Together, these technologies exemplify how AI and IoT can transform the way we observe, analyze, and understand animal life in the wild.

Biography

Takuya Maekawa is a professor at the Institute for Advanced Co-Creation Studies, The University of Osaka. In 2006, he received his doctor degree from Graduate School of Information Science and Technology, the University of Osaka. He then worked at NTT Communication Science Laboratory for six years. His research interest includes sensor-based context recognition techniques for ubiquitous/wearable computers.

Keynote 3 (Sept. 12, 9:00-10:00)
Towards Real-time Vision Perception: Rethinking Offloading and On-Device Approaches

Abstract

Real-time vision perception is emerging as a cornerstone technology across domains such as autonomous systems, augmented reality, and multi-modal AI agents. This keynote examines the current landscape and future directions of real-time vision systems by revisiting the longstanding trade-offs between cloud offloading and on-device computation. As hardware capabilities continue to advance and application demands grow more complex, challenges around latency, energy efficiency, model scalability, and privacy have become increasingly critical. We highlight recent research, including ARMA (MobiSys '25), which proposes joint scheduling of network and compute resources to guarantee real-time latency in mobile perception systems, and Logan (MobiSys '24), which rethinks video analytics by questioning whether complete packet delivery is necessary for statistical inference. These studies reveal how reimagining system-level design can unlock new possibilities for responsive and efficient vision perception. This keynote will outline key principles for building next-generation systems capable of meeting the performance demands of real-world AI applications.

Biography

Youngki Lee is an associate professor in the Department of Computer Science and Engineering at Seoul National University, where he leads the Human-Centered Computer Systems Lab (https://hcs.snu.ac.kr/). He received his Ph.D. in Computer Science from KAIST in 2013 and previously served as an assistant professor at Singapore Management University until 2018. His research focuses on designing experimental, user-aware software systems at the intersection of machine learning systems, mobile and pervasive computing, and human-computer interaction. He is particularly interested in building on-device AI systems, augmented and mixed reality platforms, and interactive AI agents for real-world applications. He actively contributes to the academic community, serving as program co-chair of ACM UbiComp 2018 and the upcoming ACM MobiSys 2026, and as a frequent technical program committee member for leading ACM SIGMOBILE conferences, including MobiSys, UbiComp, SenSys, and HotMobile.