Decoding Gaze: Mapping Human Focus in Visual Content
This presentation focuses on my research area—deep learning and computer vision applied to gaze estimation, ie. predicting a person's attention point in an image. This talk strives to provide the audience with a practical overview of this emerging field.
Emphasizing clarity, I'll introduce the task of gaze estimation and address the fundamental challenges researchers face when working on this topic. By doing so, attendees will gain insights into the intricacies of deciphering human attention in social scenes. We'll explore typical methods proposed in the literature, while showcasing the relevance of the task to real-world applications. In essence, this talk serves as an educational opportunity, offering a valuable introduction to a complex and important subject in computer vision.
Given the proficiency and diversity of the audience, I will try to make this talk as accessible as possible. That being said, some familiarity with deep learning will still be expected.
Speaker(s)
Samy Tafasca
PhD Student at the Idiap Research Institute and EPFL
Samy is a PhD student at EPFL and the Idiap Research Institute in Switzerland. Specializing in computer vision applied to human behavior understanding in social scenes, his work is driving key advancements in the field of gaze estimation. With an engineering degree from IMT Atlantique (France) and a Master’s from the University of London (UK), Samy blends academic prowess with real-world impact, having previously served as a Data Scientist in the Tech and Telco industries.
Beyond his professional pursuits, Samy is also a versatile individual. Whether immersing himself in Arabic poetry, mastering culinary arts, refining vector design skills, or exploring the globe, his interests are as diverse as the six countries he has called home.