Ihsan of University of Galway, Ireland

In this talk, Dr. Ihsan will discuss recent advances in video-based human activity recognition, focusing on two distinct contributions. The first introduces a probabilistic sampling method based on the Frobenius norm that enhances efficiency while maintaining recognition accuracy. The second presents Video-DPRP, a differentially private framework that protects visual privacy while supporting reliable activity recognition. Together, these works demonstrate the importance of both performance-driven and privacy-aware approaches in advancing video understanding, ultimately enabling more practical and trustworthy video-based activity recognition systems.

Biography:

Dr. Ihsan Ullah is a Lecturer Above the Bar in the School of Computer Science at the University of Galway, and investigator with the Data Science Institute and Insight Research Ireland center for Data Analytics. He earned his PhD from the University of Milan, where he focused on designing lightweight deep neural network architectures using pyramidal approaches. With over nine years of research and development experience, he has worked on deep learning applications for video, image, text, and time-series data in leading labs in US, EU, and Gulf. Prior to his current role, he led Special Projects at CeADAR Ireland's Centre for Applied AI, and held research/academic roles in University of Galway and UCD. His research interests include lightweight deep learning models, explainable & responsible AI, federated learning, differential privacy, and video-based human activity recognition, among others.

Email: