PROJECT TITLE :

A Reconfigurable Tangram Model for Scene Representation and Categorization

ABSTRACT:

This paper presents a hierarchical and compositional scene layout (i.e., spatial configuration) illustration and a method of learning reconfigurable model for scene categorization. Three sorts of shape primitives (i.e., triangle, parallelogram, and trapezoid), known as tans, are used to tile scene image lattice in a very hierarchical and compositional means, and a directed acyclic AND–OR graph (AOG) is proposed to arrange the overcomplete dictionary of tan instances placed in image lattice, exploring a terribly massive variety of scene layouts. With bound off-the-shelf look features used for grounding terminal-nodes (i.e., tan instances) in the AOG, a scene layout is represented by the globally optimal parse tree learned via a dynamic programming algorithm from the AOG, which we call tangram model. Then, a scene class is represented by a mix of tangram models discovered with an exemplar-based Clustering methodology. On basis of the tangram model, we address scene categorization in 2 aspects: 1) building a tangram bank representation for linear classifiers, which utilizes a collection of tangram models learned from all categories and a pair of) building a tangram matching kernel for kernel-primarily based classification, that accounts for all hidden spatial configurations in the AOG. In experiments, our strategies are evaluated on 3 scene data sets for each the configuration-level and semantic-level scene categorization, and outperform the spatial pyramid model consistently.