About me

Hello! I’m Andreu Girbau Xalabarder, a Post-Doctoral research fellow at Satoh-lab at the National Institute of Informatics in Tokyo, and my interests are in Event Cameras, Computer Vision and applied Machine Learning in general. I did my PhD at AutomaticTV|Mediapro in collaboration with the Polytechnic University of Catalonia in Barcelona with the Image Processing Group. Feel free to contact me for collaboration!

Latest News

Oct. 2023: I will be serving as Area Chair for MMM2025 in Nara, Japan.

Research interests

Event Cameras

My current research focuses on Event Cameras and their applications in computer vision. Traditional cameras capture images at a fixed frame rate while event cameras only capture changes in the scene (aka. events) per pixel as they happen. This changes the paradigm of computer vision from discrete to (almost) continuous information, opening the possibility to many different applications, such as collision avoidance in fast-paced dynamic environments.

BalloonPopping

Multiple Object Tracking

During my Ph.D., I had the privilege of working at AutomaticTV, a company dedicated to generating sports productions using computer vision techniques. Among various challenges, a significant part of my research focused on Multiple Object Tracking (MOT) and its practical applications in real-world scenarios. What made this experience really interesting was how I got to see the different ways industry and academia do things, having to tackle the MOT challenge from different angles, understanding the needs of both worlds.

TV Analysis of Political figures

I had the great opportunity to work in a multi-disciplinary project on the intersection of political science and computer vision. The objective of this project was to analyze Japanese television, with a particular focus on political trends over the last 20 years. The project involved building datasets, working in face detection, tracking, and classification, building a system as robust and flexible as possible, so we could use the system to be able to quickly analyze other political actors and other video frameworks.

GovOppJP