Visual Recognition from Large-Scaled Image Data with Limited Supervision

Lecturer: 
Prof. Eric Granger
Lecturer's institute: 
ETS Montreal, Canada
Date: 
29.10.2018 13:00
Place: 
Tellus Business Kitchen

Abstract:

State-of-the-art sensors to capture images, audio and video signals are paving the way to innovative, next-generation technologies for many applications in, e.g., video surveillance, medical diagnosis, health monitoring, content-based image/video retrieval.  For instance, the detection, tracking and recognition of actions, cars, people, etc., appearing over a distributed network of cameras are key functions for many video-based summarization and surveillance applications. Designing accurate recognition systems for these applications typically gives rise to several challenges because it involves learning complex models using large weakly-annotated data sets that incorporate domain shifts, subtle noise, variations and uncertainties. Focused on training visual recognition models from large amounts of weakly-labeled image data, this talk will discuss recent techniques for deep learning, sparse modeling, weakly-supervised learning, cross-domain adaptation, and information fusion, which are promising for complex image/video processing problems. This talk will also introduces the research activities conducted at the Laboratory of Imaging, Vision and Artificial Intelligence (LIVIA) at ETS Montreal. LIVIA's activities revolve around techniques in machine learning, computer vision, pattern recognition, adaptive and intelligent systems, information fusion, and optimization of complex systems. The application areas are: (a) the analysis of medical, satellite, aerial images, etc., (b) biometrics (recognition of individuals from the signature, the face, the voice, etc.), (c) automatic processing of handwritten documents, (d) affective computing for health monitoring, and (e) security and surveillance.

Short Biography

Prof. Eric Granger received Ph.D. in EE from École Polytechnique de Montréal in 2001, and worked as a Defense Scientist at Defense R&D Canada – Ottawa (1999-2001), and in R&D with Mitel Networks (2001-2004). He joined the École de technologie supérieure (Université du Québec), Montreal, in 2004, where he is presently Full Professor and director of LIVIA, a research laboratory focused on computer vision and artificial intelligence. His research interests include pattern recognition, machine learning, computer vision, domain adaptation, and incremental and weakly-supervised learning, with applications in biometrics, affective computing, video surveillance, and computer/network security.

 

Last updated: 19.10.2018