In this paper we propose a technique that combines a classification method from the statistical learning literature with a conventional approach to shape retrieval. The idea that we pursue is to improve both results and performance by filtering the database of shapes before retrieval with a shape classifier, which allows us to keep only the shapes belonging to the classes most similar to the query shape. The experimental analysis that we report shows that our approach improves the computational cost in the average case, and leads to better results too.

In numerous domains it is useful to represent a single example by the collection of local features or parts that comprise it. In computer vision in particular, local image features are a powerful way to describe images of objects and scenes. Their stability under variable image conditions is critical for success in a wide range of recognition and retrieval applications. However, many conventional similarity measures and machine learning algorithms assume vector inputs. Comparing and learning from images represented by sets of local features is therefore challenging, since each set may vary in cardinality and its elements lack a meaningful ordering. In this thesis I present computationally efficient techniques to handle comparisons, learning, and indexing with examples represented by sets of features. The primary goal of this research is to design and demonstrate algorithms that can effectively accommodate this useful representation in a way that scales with both the representation size as well as the number of images available for indexing or learning. I introduce the pyramid match algorithm, which efficiently forms an implicit partial matching

Abstract—In this paper a novel approach for dissimilaritybased representation is presented, which combines local image descriptors with several dissimilarity functions. The basic idea consists of defining the set of prototypes in terms of local descriptors of image parts, namely feature points extracted from the training set. Therefore, according to the dissimilarity-based approach, a new image can be characterized on the basis of its dissimilarity with each of the given prototypes. This leads to a new class of Local Kernels which exploits the use of dissimilarities between image parts. In particular, we show that the classic Bagof-Feature (BoF) kernel can be revised as a special case of our new formulation, and better performance can be obtained when new dissimilarity functions are employed. Moreover, we observe that any variants of the basic BoF kernel can take advantage from our approach as we show for the case of the Pyramid Match kernel. Promising results are shown for image categorization on the ETH-80 database. I.

We show that a class of non-linear kernel SVMs admit approximate classifiers with run-time and memory complexity that is independent of the number of support vectors. This class of kernels which we refer to as additive kernels, include the widely used kernels for histogram based image comparison like intersection and chi-squared kernels. Additive kernel SVMs can offer significant improvements in accuracy over linear SVMs on a wide variety of tasks while having the same run-time, making them practical for large scale recognition or real-time detection tasks. We present experiments on a variety of datasets including the INRIA person, Daimler-Chrysler pedestrians, UIUC Cars, Caltech-101, MNIST and USPS digits, to demonstrate the effectiveness of our method for efficient evaluation of SVMs with additive kernels. Since its introduction, our method has become integral to various state of the art systems for PASCAL VOC object detection/image classification, ImageNet Challenge, TRECVID, etc. The techniques we propose can also be applied to settings where evaluation of weighted additive kernels is required, which include kernelized versions of PCA, LDA, regression, k-means, as well as speeding up the inner loop of SVM classifier training algorithms.