Speaker: Prof. Benjamin Risse

Abstract: The predicted market for AI-based computer vision solutions indicates exponential growth over the next 10 years, reflecting the immense expectations for upcoming improvements in this area. Interestingly, despite the immense progress in the fundamental research of machine learning in general, and computer vision in particular, past projections have usually overestimated the applicability of these technologies by a large margin. The goal of my presentation is to discuss potential reasons for these shortcomings while providing an alternative approach, which integrates applicability by design while still focusing on the theory of the underlying algorithms. Based on several research examples, I will present how the latest deep learning models can be integrated into emerging technologies, such as event cameras, eye tracking, and extended reality, to yield novel approaches and tools for a variety of scalable use cases and real-world applications.

Short Bio: Since 2018, Benjamin Risse has been a professor at the University of Münster and the head of the Computer Vision & Machine Learning Systems (CVMLS) group. He studied computer science and received a PhD at the intersection of computer vision, machine learning, and neuroscience. From 2015 to 2017, he was a postdoctoral researcher at the University of Edinburgh. The research of the CVMLS group focusses on interdisciplinary computer vision and machine learning questions with a particular interest in the usage of this technology in the context of wildlife conservation, movement quantifications and the applications of AI in virtual and augmented reality applications.