Kernel-based similarity learning

In content-based image retrieval (CBIR), a set of low-level features are extracted from an image to represent its visual content. Retrieval is performed by image example where a query image is given as input by the user and an appropriate similarity measure is used to find the best matches in the corresponding feature space. This approach suffers from the fact that there is a large discrepancy between the low-level visual features that one can extract from an image and the semantic interpretation of the image’s content that a particular user may have in a given situation. That is, users seek semantic similarity, but we can only provide similarity based on low-level visual features extracted from the raw pixel data, a situation known as the semantic gap. The selection of an appropriate similarity measure is thus an important problem. Since visual content can be represented by different attributes, the combination and importance of each set of features varies according to the user’s semantic intent. Thus, the retrieval strategy should be adaptive so that it can accommodate the preferences of different users. Relevance feedback (RF) learning has been proposed as a technique aimed at reducing the semantic gap. It works by gathering semantic information from user interaction. Based on the user’s feedback on the retrieval results, the retrieval scheme is adjusted. By providing an image similarity measure under human perception, RF learning can be seen as a form of supervised learning that finds relations between high-level semantic interpretations and low-level visual properties. That is, the feedback obtained within a single query session is used to personalize the retrieval strategy and thus enhance retrieval performance. In this chapter we present an overview of CBIR and related work on RF learning. We also present our own previous work on a RF learning-based probabilistic region relevance learning algorithm for automatically estimating the importance of each region in an image based on the user’s semantic intent.

In a content-based image retrieval (CBIR) system, indexing feature vectors and the similarity measure between feature vectors are two key factors for retrieval performance. We present a new CBIR system with statistical-model based image feature extraction in the wavelet domain and a Kullback divergence based similarity measure. A two component Gaussian mixture model (GMM) in the wavelet domain is employed and the model parameters are used to form features for image indexing. A new Kullback divergence based similarity measure is then presented for image retrieval. The experimental results demonstrate that the similarity measure based on the Kullback divergence is more effective than conventional similarity measures, such as the city-block distance and the Euclidean distance. It is shown that the new CBIR system, with the combination of the GMM and the new Kullback divergence based similarity measure, outperforms most other methods in retrieval performance for texture images, while keeping a comparable level of computational complexity.

In the current years, advancements in medical imaging have prompted the rise of some massive databases, including the pictures from a different scope of modalities. Content based image retrieval (CBIR) is a kind of search that selects examples from an image set that have a similar content to an input query image. Each CBIR system consists of two substantial factors, feature extraction and similarity measure. Based on the type of these parts, many CBIR systems are proposed. In this paper we present not only how to design a CBIR system for MRI brain Tumor images, but also a comparison between the different implemented CBIR systems and ours. Furthermore, in this research, using a DML (Distance Metric Learning) instead of the rigid distance metric such as Euclidean distance as the Similarity measure has helped the researchers to obtain a better mAP (mean Average Precision) as the result. The experimental results of our CBIR system show that it has a better performance than other rivals in the field. The obtained mean average precision of our CBIR system is 92.41 that is significant in MRI brain tumor.

In this paper, we present a novel approach for medical content based image retrieval (searching for images which are pathologically similar to a given example) and demonstrate its performance on a dataset containing maxillofacial lesions. Recently, distributed databases at hospitals and CT scanning centers are related through Picture Archiving and Communication Systems (PACS). Hence, a content based image retrieval system is helpful for radiologists in medical diagnosis. In our proposed framework, a feature vector is extracted for each image and SIFT sparse codes are employed in this step. After feature extraction based on sparse coding and maximum pooling, we have utilized different similarity measures such as Euclidean norm, Manhattan distance and SVM classifier to choose most relevant images to the query image. We have evaluated our proposed framework on a dataset containing 2023 images belonging to 5 different categories: cleft, impaction, fracture, maxillary sinus cyst and healthy. Classification rate of 97.3% and precision versus recall curve indicate the effectiveness of sparse coding in content-based medical image retrieval.

A challenging problem of image retrieval is the similarity learning between images. To improve similarity search in Content-Based Image Retrieval (CBIR), many studies on distance metric learning have been published. Despite their success, most existing methods are limited in two aspects: (i) they usually attempt to learn a linear distance metric, which limits their capacity of measuring similarity for complex applications; (ii) they are often designed for learning metrics on unique-modal data, which could be suboptimal for similarity learning on multimedia objects with multiple feature representations. To overcome these limitations, in this paper, we investigate the online kernel-based multimodal similarity learning, which aims to integrate multiple kernels for learning multimodal similarity functions, and conduct experiments to evaluate the performance of the proposed method for CBIR on several different image datasets. The experiment results are encouraging and verify the effectiveness and the superiority of the proposed method.

Practical content-based image retrieval systems require efficient relevance feedback techniques. Researchers have proposed many relevance feedback methods using quadratic-form distance metric as similarity measure and learning similarity matrix by feedback samples. Existing methods fail to find the optimal and reasonable solution of similarity measure due to the small number of positive and negative training samples. In this paper, an approach of learning the similarity measure using feature subspaces analysis (FSA) is proposed for content-based image retrieval. This approach solves the similarity measure-learning problem by FSA on training samples, which improves generalization capacity and reserves robustness furthest simultaneously. Experiments on a large database of 13,897 heterogeneous images demonstrated a remarkable improvement of retrieval precision.

Content-based image retrieval (CBIR) has turned into an important and active potential research field with the advance of multimedia and imaging technology. It makes use of image features, such as color, texture and shape, to index images with minimal human intervention. A CBIR system can be used to locate medical images in large databases. In this paper we propose a CBIR system which describes the methodology for retrieving digital human brain magnetic resonance images (MRI) based on textural features and the Adaptive neuro-fuzzy inference system (ANFIS) learning to retrieve similar images from database in two categories: normal and tumoral. A fuzzy classifier has been used, because of the uncertainty in the results of classifier and capacity of learning. ANFIS is a good candidate for our categorization problem. Our proposed CBIR system can locate a query image in the category of normal or tumoral images in the online retrieval part. Finally, using a relevance feedback, we improve the effectiveness of our retrieval system. This research uses the knowledge of the CBIR approach to the application of medical decision support and discrimination between the normal and abnormal medical images based on features. We present and compare the results of the proposed method with the CBIR systems used in recent works. The experimental results indicate that the proposed method is reliable and has high image retrieval efficiency compared with the previous works.

Content Based Image Retrieval (CBIR) is a process in which for a given query image similar images will be retrieved based on the image content similarity. Image content refers to its visual features, which are mathematical representations of a digital image. The image retrieval task primarily depends on image feature extraction and similarity measurement between the feature vectors. The performance of CBIR process not only decided by the optimum features extracted from the image but also on the proper choice of the similarity/dissimilarity measures (distance metrics). As the image features broadly diversified in terms of color, texture and shape based, using the same distance metric may not work well for all of these features. In this paper, first we presented overview of geometric and statistical distance metrics used in CBIR along with the comparative analysis of these measures on color and texture features. Color features extracted by computing color histograms in HSV space and texture features by wavelet decompositions. Geometrical distances such as Manhattan, Chebyshev, Euclidean, statistical distance metrics such as Cosine Similarity, Chi-square, KullbackLeibler, Jeffrey and cumulative statistical distance metrics of Kolmogorov-Smirnov, Cramer von Mises and Earthmover's distances were analyzed for feature similarity. We gave certain conclusions on the performance of all these distance metrics in terms of Mean Average Precision (MAP) and Recall rates with color and texture features respectively.

With the emergence and explosion of huge image databases there is an increasing necessity for effective methods to assess visual information on the level of objects and scene types. A wide variety of Content – Based Image Retrieval (CBIR) systems already exists. As a key issue in CBIR, similarity measure quantifies the resemblance in contents between a pair of images. Depending on the type of features, the formulation of the similarity measure varies greatly. The primary goal of our study is to reduce the computation time and user interaction. The secondary goal is to reduce the semantic gap between high level concepts and low level features. A third goal is to evaluate system performance with regard to speed and accuracy. In the proposed study transform color after statistical transform, such as the Adaptive Color Karhunen Loeve Transform (ACKLT) is used as a color descriptor. The results are showing the advantage of the new algorithm for ACKLT in comparison with the YCrCb color model. Based on the experimental results, we concluded that correct selection of descriptors invariant to light intensity and light color changes affects object and scene category recognition.Keywordscontent-based image retrievalAdaptive Color Karhunen Loeve Transform

The performance of content-based image retrieval (CBIR) is usually limited since only single visual feature and single similarity measurement are used. In order to solve this problem, the color and texture visual features of an image are analyzed firstly. And then 12 kinds of similarity measurement are used to evaluate similarity between the image being checked and the images in the retrieval library. The CBIR problem is therefore transferred to an optimization problem with the precision ratio as its objective function. Quantum-behaved Particle Swarm Optimization (QPSO) algorithm is used to solve the CBIR optimization problem in order to find the optimal weight and the optimal combination of visual features and similarity measurements. Experimental results show that the proposed method based on QPSO algorithm has better performance on the retrieval effect.

This paper presents an efficient image retrieval system called BICA-C for face recognition. This system takes an image as an input query and retrieves images based on its image content. Content Based Image Retrieval (CBIR) is an approach for retrieving semantically-relevant images from an image database based on extracted image features. Block Independent Component Analysis (B-ICA) is a popular feature extraction method. This paper proposes a scheme to enhance the performance of B-ICA based feature extraction by the utilization of k-means clustering technique. The proposed scheme consists of two stages. First stage is the offline training process where the extracted features in the database is clustered using k-means clustering and then the second stage is the online testing where the obtained query image is compared with the clustered images and is classified as intruder or authenticated person based on the similarity measure.

A new scheme of learning similarity measure is proposed for content-based image retrieval (CBIR). It learns a boundary that separates the images in the database into two parts. Images on the positive side of the boundary are ranked by their Euclidean distances to the query. The scheme is called restricted similarity measure (RSM), which not only takes into consideration the perceptual similarity between images, but also significantly improves the retrieval performance based on the Euclidean distance measure. Two techniques, support vector machine and AdaBoost, are utilized to learn the boundary, and compared with respect to their performance in boundary learning. The positive and negative examples used to learn the boundary are provided by the user with relevance feedback. The RSM metric is evaluated on a large database of 10,009 natural images with an accurate ground truth. Experimental results demonstrate the usefulness and effectiveness of the proposed similarity measure for image retrieval.

Although rare in the sound recognition literature, previous work using features derived from a sparse temporal representation has led to some success [8, 2, 9]. A great advantage of deriving features from a temporal representation is that such an approach does not face the trade-off problem between time and frequency resolution. Here, we present a biologically inspired two-step process for audio classification: In the first step, efficient basis functions are learned in an unsupervised manner [12] on mixtures of percussion sounds (drum phrases). In the second step, features are extracted by using the learned basis functions to decompose percussion sounds (base drum, snare drum, hi-hat) with matching pursuit [7]. The classification accuracy in a 3-class database transfer task is 91.5% as opposed to 70.7% when using MFCC features. Further, we show that a MP-feature representation preserves sound similarity to a greater extent than MFCC-features, i.e. an artificial mixture of two sounds of equal energy normally lies in the middle between the two single sound distributions in feature space. An MP-representation thus inherently contains a similarity measure between different sounds.

Defining similarity measures is a requirement for some machine learning methods. One such method is case-based reasoning (CBR) where the similarity measure is used to retrieve the stored case or a set of cases most similar to the query case. Describing a similarity measure analytically is challenging, even for domain experts working with CBR experts. However, datasets are typically gathered as part of constructing a CBR or machine learning system. These datasets are assumed to contain the features that correctly identify the solution from the problem features; thus, they may also contain the knowledge to construct or learn such a similarity measure. The main motivation for this work is to automate the construction of similarity measures using machine learning. Additionally, we would like to do this while keeping training time as low as possible. Working toward this, our objective is to investigate how to apply machine learning to effectively learn a similarity measure. Such a learned similarity measure could be used for CBR systems, but also for clustering data in semi-supervised learning, or one-shot learning tasks. Recent work has advanced toward this goal which relies on either very long training times or manually modeling parts of the similarity measure. We created a framework to help us analyze the current methods for learning similarity measures. This analysis resulted in two novel similarity measure designs: The first design uses a pre-trained classifier as basis for a similarity measure, and the second design uses as little modeling as possible while learning the similarity measure from data and keeping training time low. Both similarity measures were evaluated on 14 different datasets. The evaluation shows that using a classifier as basis for a similarity measure gives state-of-the-art performance. Finally, the evaluation shows that our fully data-driven similarity measure design outperforms state-of-the-art methods while keeping training time low.

Content Based Image Retrieval (CBIR) is important in computer aided plant species recognition. Texture and shape information have been the primitive image descriptors in content based image retrieval systems. This paper presents a novel framework for combining both the features, texture and shape information, and achieve higher retrieval efficiency. The study provides a methodology for retrieving medicinal plants images from a database of medicinal plant images based on shape and texture features. The shape descriptors include Zernike moment, Fourier descriptor (FD), Generic fourier Descriptor(GFD) and for texture descriptors gabor filters are used. The similarity measures, euclidean distance of each medicinal plant image from the database to query image is used. The images are sorted based on similarity of Euclidean distance. The retrieval experiments are carried on different training and test medicinal plant images. The effectiveness of different descriptors is confirmed by the experimental results. We have investigated shape and texture features for medicinal plant retrieval by successively combining the different transforms. The retrieval efficiency is reported through precision and recall rate. Experimental results by combining Gabor and Zernike transform outperforms the all other methods.