Shape Matching and Object Recognition Using Low Distortion Correspondences

TLDR

The authors aim to perform object recognition by deformable shape matching, introducing a new algorithm that finds correspondences between feature points. The algorithm formulates correspondence as an integer quadratic program with a cost based on geometric blur descriptors and distortion, handles outliers, estimates a regularized thin‑plate spline transform, and uses the resulting matching cost as a nearest‑neighbor distance for recognition. On Caltech 101 the method attains 48 % classification accuracy versus 16 % by Fei‑Fei et al., and it localizes frontal and profile faces with performance comparable to specialized face‑tuned approaches.

Abstract

We approach recognition in the framework of deformable shape matching, relying on a new algorithm for finding correspondences between feature points. This algorithm sets up correspondence as an integer quadratic programming problem, where the cost function has terms based on similarity of corresponding geometric blur point descriptors as well as the geometric distortion between pairs of corresponding feature points. The algorithm handles outliers, and thus enables matching of exemplars to query images in the presence of occlusion and clutter. Given the correspondences, we estimate an aligning transform, typically a regularized thin plate spline, resulting in a dense correspondence between the two shapes. Object recognition is then handled in a nearest neighbor framework where the distance between exemplar and query is the matching cost between corresponding points. We show results on two datasets. One is the Caltech 101 dataset (Fei-Fei, Fergus and Perona), an extremely challenging dataset with large intraclass variation. Our approach yields a 48% correct classification rate, compared to Fei-Fei et al 's 16%. We also show results for localizing frontal and profile faces that are comparable to special purpose approaches tuned to faces.