Large-scale image categorization with explicit data embedding

Perronnin, Florent; Senchez, Jorge; Xerox, Yan Liu

doi:10.1109/cvpr.2010.5539914

Cited by 121 publications

(91 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Actually, the square-rooting BoW histogram is the feature map of Hellinger kernel k x; y ð Þ ¼ ffiffiffiffi ffi xy p , which can improve the retrieval performance [33,34]. To develop the CBIR system of contrast-enhanced liver CT images, the visual vocabularies of each phase and the distance metrics were learned offline for each single-or multiphase CT image.…”

Section: Bag-of-visual-words Representation Of Lesionsmentioning

confidence: 99%

Content-Based Retrieval of Focal Liver Lesions Using Bag-of-Visual-Words Representations of Single- and Multiphase Contrast-Enhanced CT Images

Yang

et al. 2012

J Digit Imaging

View full text Add to dashboard Cite

This paper is aimed at developing and evaluating a content-based retrieval method for contrastenhanced liver computed tomographic (CT) images using bag-of-visual-words (BoW) representations of single and multiple phases. The BoW histograms are extracted using the raw intensity as local patch descriptor for each enhance phase by densely sampling the image patches within the liver lesion regions. The distance metric learning algorithms are employed to obtain the semantic similarity on the Hellinger kernel feature map of the BoW histograms. The different visual vocabularies for BoW and learned distance metrics are evaluated in a contrast-enhanced CT image dataset comprised of 189 patients with three types of focal liver lesions, including 87 hepatomas, 62 cysts, and 60 hemangiomas. For each single enhance phase, the mean of average precision (mAP) of BoW representations for retrieval can reach above 90 % which is significantly higher than that of intensity histogram and Gabor filters. Furthermore, the combined BoW representations of the three enhance phases can improve mAP to 94.5 %. These preliminary results demonstrate that the BoW representation is effective and feasible for retrieval of liver lesions in contrast-enhanced CT images.

show abstract

Section: Bag-of-visual-words Representation Of Lesionsmentioning

confidence: 99%

Content-Based Retrieval of Focal Liver Lesions Using Bag-of-Visual-Words Representations of Single- and Multiphase Contrast-Enhanced CT Images

Yang

et al. 2012

J Digit Imaging

View full text Add to dashboard Cite

show abstract

“…For smallscale datasets, it is feasible to learn a non-linear SVM, albeit the training complexity is somewhere between quadratic and cubic [20]. In general, however, nonlinear SVMs do not scale well with training set size.…”

Section: Data Embeddingmentioning

confidence: 99%

“…Since the semantic kernel of (4) is not additive, neither the embedding of [19], nor the embedding learning method of [20] are feasible. One alternative, that we explore, is to replace the arccos term by a first-order Taylor series around 0, i.e., arccos…”

Section: Data Embeddingmentioning

confidence: 99%

“…Although the semantic space representation is much more compact than BoW, its combination with spatial encoding mechanisms and large theme vocabularies can also lead to large-scale learning problems. While in the BoW literature some authors have proposed explicit data embedding strategies [19,20], which enable the replacement of non-linear by linear SVMs, greatly reducing computation, such embeddings are not yet available for theme-based representations.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Scene Recognition on the Semantic Manifold

Kwitt

Vasconcelos

Rasiwasia³

2012

Computer Vision – ECCV 2012

View full text Add to dashboard Cite

Abstract. A new architecture, denoted spatial pyramid matching on the semantic manifold (SPMSM), is proposed for scene recognition. SPMSM is based on a recent image representation on a semantic probability simplex, which is now augmented with a rough encoding of spatial information. A connection between the semantic simplex and a Riemmanian manifold is established, so as to equip the architecture with a similarity measure that respects the manifold structure of the semantic space. It is then argued that the closed-form geodesic distance between two manifold points is a natural measure of similarity between images. This leads to a conditionally positive definite kernel that can be used with any SVM classifier. An approximation of the geodesic distance reveals connections to the well-known Bhattacharyya kernel, and is explored to derive an explicit feature embedding for this kernel, by simple squarerooting. This enables a low-complexity SVM implementation, using a linear SVM on the embedded features. Several experiments are reported, comparing SPMSM to state-of-the-art recognition methods. SPMSM is shown to achieve the best recognition rates in the literature for two large datasets (MIT Indoor and SUN) and rates equivalent or superior to the state-of-the-art on a number of smaller datasets. In all cases, the resulting SVM also has much smaller dimensionality and requires much fewer support vectors than previous classifiers. This guarantees much smaller complexity and suggests improved generalization beyond the datasets considered.

show abstract

“…However, it was recently shown that for some particular classes of kernels such approximations do exist. The main lines of research towards efficient approximation of features map include (a) exploiting particular kernel properties to find the approximation (e.g., in [27,18] the authors exploited various properties to propose efficient and effective approximations of large families of additive kernels); (b) the application of random sampling on Fourier features [15,16,2] (e.g., in [22] methodologies have been proposed for encoding stationary kernels by randomly sampling their Fourier features); (c) the application of the so-called Nystrom method, which is a data-dependent methodology that requires training [29,28,21]. Even though the above methods provide useful and general methodologies that are applicable to many kernels, their disadvantage is that they provide approximate solutions.…”

Section: Introductionmentioning

confidence: 99%

On One-Shot Similarity Kernels: Explicit Feature Maps and Properties

Zafeiriou

Kotsia

2013

2013 IEEE International Conference on Computer Vision

View full text Add to dashboard Cite

Kernels have been a common tool of machine learning and computer vision applications for modeling nonlinearities and/or the design of robust 1 similarity measures between objects. Arguably, the class of positive semidefinite (psd) kernels, widely known as Mercer's Kernels, constitutes one of the most well-studied cases. For every psd kernel there exists an associated feature map to an arbitrary dimensional Hilbert space H, the so-called feature space. The main reason behind psd kernels' popularity is the fact that classification/regression techniques (such as Support Vector Machines (SVMs)) and component analysis algorithms (such as Kernel Principal Component Analysis (KPCA)) can be devised in H, without an explicit definition of the feature map, only by using the kernel (the so-called kernel trick). Recently, due to the development of very efficient solutions for large scale linear SVMs and for incremental linear component analysis, the research towards finding feature map approximations for classes of kernels has attracted significant interest. In this paper, we attempt the derivation of explicit feature maps of a recently proposed class of kernels, the so-called one-shot similarity kernels. We show that for this class of kernels either there exists an explicit representation in feature space or the kernel can be expressed in such a form that allows for exact incremental learning. We theoretically explore the properties of these kernels and show how these kernels can be used for the development of robust visual tracking, recognition and deformable fitting algorithms.1 Robustness may refer to either the presence of outliers and noise or to the robustness to a class of transformations (e.g., translation).

show abstract

Large-scale image categorization with explicit data embedding

Cited by 121 publications

References 12 publications

Content-Based Retrieval of Focal Liver Lesions Using Bag-of-Visual-Words Representations of Single- and Multiphase Contrast-Enhanced CT Images

Content-Based Retrieval of Focal Liver Lesions Using Bag-of-Visual-Words Representations of Single- and Multiphase Contrast-Enhanced CT Images

Scene Recognition on the Semantic Manifold

On One-Shot Similarity Kernels: Explicit Feature Maps and Properties

Contact Info

Product

Resources

About