Spatiotemporal Contour Grouping Using Abstract Part Models

Part Models Pablo Sala1 , Diego Macrini2 , and Sven Dickinson1 1 University of Toronto, 2 Queen's University Abstract. In recent work [1], we introduced a framework for modelrceptual grouping and shape abstraction using a vocabulary of art shapes. Given a user-defined vocabulary of simple abstract he framework grouped image contours whose abstract shape was consistent with one of the part models. While the results showed promise, the representational gap between the actual image contours that make emplar shape and the contours that make up an abstract part significant, and an abstraction of a group of image contours may be consistent with more than one part model; therefore, while recall of ground-truth parts was good, precision was poor. In this paper, we address the precision problem by moving the camera and exploiting spatiotemporal constraints in the grouping process. We introduce a novel probabilistic, graph-theoretic formulation of the problem, in which the spatiotemporal co...

• Type: Journal
• Year: 2010
• Where: ACCV
• Authors: Pablo Sala, Diego Macrini, Sven J. Dickinson
• Keywords: ACCV 2010, Computer Vision, Image Contours, Spatiotemporal Constraints, Static Image Version