Informatics, TU Vienna

Interactive Image Segmentation: Research & Product

Interactive image segmentation is the task of extracting an object from an image with as few user interactions as possible.

Der Arbeitsbereich für Interaktive Mediensysteme lädt gemeinsam mit dem Austrian Institute of Technology (AIT) und der IEEE Austria Section ein.
 

Abstract

Interactive image segmentation is the task of extracting an object from an image with as few user interactions as possible. A technology for this, called GrabCut, will ship with Microsoft Office 2010. In this talk I will describe a series of research work which inspired the final product. In particular, I will concentrate on the research question of jointly optimizing a segmentation functional together with an unknown appearance model. Although this is in general an NP-hard problem, I will explain several practical algorithms, which can also be used in other domains such as recognition and segmentation. Another aspect of this talk is how to integrate the user interactions into the optimization process. In this way, I will give the key ideas of three recent articles from ICCV and ECCV conferences.

Biography 

Since 2004, Carsten Rother is a permanent researcher at Microsoft Research Cambridge. He received a PhD from the Royal Institute of Technology inStockholm, Sweden. His research interests are in the field of "(Markov) Random Field Models for Computer Vision", "Discrete Optimization", and
"Vision for Graphics" (in particular interactive segmentation and matting).
He has published more than 20 high impact papers (at least 10 citations) at international conferences and journals, and co-authored a paper which won the best paper honourable mention award at CVPR'05. He serves on the program committee of major conferences (e.g. SIGGRAPH, ICCV, ECCV, CVPR, NIPS), has been area chair for BMVC'08, and '09, and he has organized workshops at ICCV'07 and ICCV'09.