Improving Shape Retrieval by Learning Graph Transduction

Abstract

Shape retrieval/matching is a very important topic in computer vision. The recent progress in this domain has been mostly driven by designing smart features for providing better similarity measure between pairs of shapes. In this paper, we provide a new perspective to this problem by considering the existing shapes as a group, and study their similarity measures to the query shape in a graph structure. Our method is general and can be built on top of any existing shape matching algorithms. It learns a better metric through graph transduction by propagating the model through existing shapes, in a way similar to computing geodesics in shape manifold. However, the proposed method does not require learning the shape manifold explicitly and it does not require knowing any class labels of existing shapes. The presented experimental results demonstrate that the proposed approach yields significant improvements over the state-of-art shape matching algorithms. We obtained a retrieval rate of 91% on the MPEG-7 data set, which is the highest ever reported in the literature.

Dynamic Label Propagation for Semi-supervised Multi-class Multi-label Classification

“…Note that µ and σ are hyper-parameters. σ is learned by the mean distance to K-nearest neighborhoods [31]. A natural transition matrix on V can be defined by normalizing the weight matrix as:…”

Balancing Deformability and Discriminability for Shape Matching

“…Recent work clearly demonstrated that adding context information to direct pairwise shape similarity can substantially improver shape retrieval [40,15,41]. Under context of a given shape we understand here its first K nearest neighbors.…”

“…Under context of a given shape we understand here its first K nearest neighbors. However, these methods [40,15,41] mainly focus on improving the transduction algorithms. We demonstrate that a 'better' original distance matrix is also very crucial for the shape retrieval with context information.…”

Exploiting Contextual Information for Image Re-ranking

“…Some post-processing methods have been proposed for improving effectiveness of information retrieval tasks [5,18,17,9]. Efforts were put on post-processing the similarity scores by analyzing the relations among all documents in a given collection.…”

“…An unsupervised clustering algorithm is used in [5], aiming to capture the manifold structure of the image relations by defining a neighborhood for each data point in terms of a mutual k-nearest neighbor graph. A graph transduction learning approach is introduced in [17]. The algorithm computes the shape similarity of a pair of shapes in the context of other shapes.…”