Pioneer Centre for AI Talk: Sergio Escalera

b

Title

Video Transformers

Abstract

In this talk professor Sergio Escalera will introduce his computer vision lab in Barcelona and some of the topics they are working on. A student of Sergio’s, Javier Selva, will then take over and present a deep dive into one of the topics, namely video transformers. As transformers see novel adaptations to many modalities, showing impressive results, it may be difficult to keep track. For this reason, he will present an overview of video transformers, delving into recent developments, highlighting main trends of designing and training video transformers, as well as some analysis of performance on classification.

Bio

Sergio Escalera is Full Professor at the Department of Mathematics and Informatics, Universitat de Barcelona. He leads the Human Behavior Analysis Group. He is Distinguished Professor at Aalborg University. He is also a member of the Computer Vision Center. He is vice-president of ChaLearn Challenges in Machine Learning, leading ChaLearn Looking at People events. He is co-creator of Codalab open source platform for challenges organization and co-founder of the NeurIPS competition and Datasets & Benchmarks tracks. He is also Fellow of the ELLIS European Laboratory for Learning and Intelligent Systems working within the Human-centric Machine Learning program, Fellow of the International Association for Pattern Recognition and vice-Chair of IAPR TC-12: Multimedia and visual information systems, Fellow of AAIA Asia-Pacific Artificial Intelligence Association, member of the AAAC Association for the Advancement of Affective Computing, and Senior IEEE member. He participated in several international funded projects and received an Amazon Research Award. He has published more than 400 research papers and participated in the organization of scientific events. He received a CVPR best paper award nominee and a CVPR outstanding reviewer award. His research interests include inclusive and transparent analysis of humans from visual and multi-modal data.

 
Javier Selva with a bachelor of Computer Science and a MSc of Artificial Intelligence, he is currently pursuing a PhD on Computer Vision under Prof. Sergio Escalera's supervision. In particular, he focuses on learning video representations in the context of human analysis. He is mainly interested in self-supervised learning and interpretable models. He is currently researching the state of the art of video transformers and self-supervised contrastive methods for video.