Distance metric for motion vector histograms based on human perceptual characteristics

The similarity metric between videos is integral to several key tasks, including video retrieval, classification and recommendation. Since there is no standard criterion for the similarity measurement between videos except measuring manually, it is difficult to collect large training dataset for distance metric learning algorithms. Moreover, the existing distance metric learning (DML) methods for multimedia data suffer from two critical limitations: (1) they typically attempt to learn a distance function on the single label setting, in which each item is only labeled with single label; (2) they are often designed for learning distance metrics on low-level features, which ignore the semantic similarity of the multimedia data. To address these problems, in this paper, we propose a novel framework of Intermediate Semantics based Distance Learning (ISDL) for video clips, which aims to integrate semantics of multiple modals optimally for distance metric learning. In particular, the proposed framework: (1) generates the training pairs automatically; (2) defines multi-modal concepts for similarity measure among videos; (3) learns the distance metric for video clips based on the intermediate semantics. We conduct an extensive set of experiments to evaluate the performance of the proposed algorithms, and the results validate the effectiveness of our proposed approach.

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.

Distance metric forms the basis of pattern classification, as almost all classifiers depend on such a metric for making classification decisions. However, the existing research testifies that a single‐distance metric is not robust enough for classification. In this Letter, the authors propose a dual distance metrics method and modify collaborative representation using dual distance metrics, that is, collaborative representation related class distance and conventional sample distance. These two distance metrics are fused by a parameter‐free multiplication scheme. The rationale of the designed multiplication fusion can be interpreted from the viewpoint of probability. As the multiplication fusion exploit the two distance metrics effectively and allow the best of the distance metrics to dominate the final classification decision, the improved collaborative representation based on dual distance metric can achieve a very high accuracy rate. Experimental results show that the proposed method has promising performance. It is also noted that the values of the two distance metrics vary with the test sample, therefore, the two distance metrics play different roles for different test samples. In other words, the two distance metrics are adaptive for test samples. The idea of the dual distance metrics method is also suitable for other combinations of other distance metrics.

Similarity measurement plays an important role to solve many pattern recognition problems such as classification, clustering and particularly the content based image retrieval problems. In this paper, a detailed comparison of 10 similarity measures has been presented for medical image retrieval. Different types of medical images from databases such as NEMA, OASIS and EXACT09 are used to evaluate the performance. Features of all the medical images are extracted using Log-Gabor wavelet. The effectiveness of all the distance metrics are investigated and tested on more than 800 medical images from the databases. It is observed from the experimental results that the retrieval performance can be improved by the distance measures like Bray- Curtis and Canberra distance metric as compared to other existing distance based approach.

This paper represents a new approach for face recognition that incorporates Prewitt edge detection, Gabor filter and Zernike moments to transform the image into a unified domain. On this joint domain, five distance metrics are constructed using Schoenberg transform for the purpose of defining efficient similarity measures for holistic face recognition. The proposed Schoenberg similarity applies Schoenberg transform to the logarithm of five existing distance metrics: Minkowski, City-Block, Euclidean, Soergel and Lorentzian metrics. The constructed Schoenberg logarithmic metrics are called SL-Minkowski, SL-City-Block, SL-Euclidean, SL-Soergel and SL-Lorentzian. These distance metrics are utilized as similarity measures after being normalized over the range [0,1] for fair comparison with existing measures. The proposed Schoenberg system can resist three problems: Change in illumination, pose and facial expression. Simulation results show that the proposed distance measures have superior performance as compared to the classical metrics: Structural Similarity Index Measure (SSIM) and Feature-based Similarity Measure (FSIM). Performance criteria are the recognition rate and the recognition confidence, defined as the similarity difference between the best match and the second-best match in the face database.

We investigate the feasibility of constructing a data-driven distance metric for use in null-hypothesis testing in the context of arms-control treaty verification. The distance metric is used in testing the hypothesis that the available data are representative of a certain object or otherwise, as opposed to binary-classification tasks studied previously. The metric, being of strictly quadratic form, is essentially computed using projections of the data onto a set of optimal vectors. These projections can be accumulated in list mode. The relatively low number of projections hampers the possible reconstruction of the object and subsequently the access to sensitive information. The projection vectors that channelize the data are optimal in capturing the Mahalanobis squared distance of the data associated with a given object under varying nuisance parameters. The vectors are also chosen such that the resulting metric is insensitive to the difference between the trusted object and another object that is deemed to contain sensitive information. Data used in this study were generated using the GEANT4 toolkit to model gamma transport using a Monte Carlo method. For numerical illustration, the methodology is applied to synthetic data obtained using custom models for plutonium inspection objects. The resulting metric based on a relatively low number of channels shows moderate agreement with the Mahalanobis distance metric for the trusted object but enabling a capability to obscure sensitive information.

Distance metric is widely used in similarity estimation. The smaller the distance is, the greater the similarity is. The Minkowski distance metric are usually chosen as the similarity measure in the conventional similarity metrics. In this paper, the most popular Euclidean and Manhattan distance, and the morphology similarity distance we have proposed are compared as similarity estimation measure distance. We cluster thirty random datasets using the fuzzy c-mean algorithm, recognize the Iris data from the UCI repository, and the experiment results are compared and analyzed.

Multi-manifold metric learning for face recognition based on image sets

Distance metric learning has been of wide concern for its prominent influence to many methods in machine learning and pattern recognition . The earlier metric learning algorithms mostly focus on finding distance metric from sample vectors. However, when dealing with images, the existing approaches often encounter high computational cost due to huge dimensionality of image vector space and instability in performance caused by distortion effect of image vectorization. This paper proposes a new technique coined Local Image Distance Metric Learning(LIDML) for image distance measurement. LIDML learns an appropriate distance metric for a novel image distance by optimizing local compactness and local separability. In LIDML, the size of distance metric is comparatively small and the model processes a closed-form solution via eigen-decomposition, both of which make LIDML fairly efficient. The experimental results on image classification demonstrate effectiveness of our method, and also show that LIDML can handle expensive computation difficulty case that hardly be handled satisfactorily by previous methods.

In this paper, we propose a method for matching biometric data from disparate domains. Specifically, we focus on the problem of comparing a low-resolution (LR) image with a high-resolution (HR) one. Existing coupled mapping methods do not fully exploit the HR information or they do not simultaneously use samples from both domains during training. To this end, we propose a method that learns coupled distance metrics in two steps. In addition, we propose to jointly learn two semi-coupled bases that yield optimal representations. In particular, the HR images are used to learn a basis and a distance metric that result in increased class-separation. The LR images are then used to learn a basis and a distance metric that map the LR data to their class-discriminated HR pairs. Finally, the two distance metrics are refined to simultaneously enhance the class-separation of both HR class-discriminated and LR projected images. We illustrate that different distance metric learning approaches can be employed in conjunction with our framework. Experimental results on Multi-PIE and SCface, along with the relevant hypothesis tests, provide evidence of the effectiveness of the proposed approach.

Comparative analysis of k-nearest neighbors distance metrics for retrieving coastal water quality based on concurrent in situ and satellite observations.

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.

Context. The development of effective distance metrics and similarity measures for categorical features is an important task in data analysis, machine learning, and decision theory since a significant portion of object properties is described by non-numerical values. Typically, the dependence between categorical features may be more complex than simply comparing them for equality or inequality. Such attributes can be relatively similar, and to construct an effective model, it is necessary to consider this similarity when calculating distance or similarity measures.
 Objective. The aim of the study is to improve the efficiency of solving practical data analysis problems by developing mathematical tools for determining the similarity of objects based on categorical ordered features.
 Method. A distance based on weighted Manhattan distance and a similarity measure for determining the similarity of objects based on categorical ordinal features (i.e. a linear order with scales of preference considering the problem domain can be specified on the attribute value set) are proposed. It is proven that the distance formula satisfies the axioms of non-negativity, symmetry, triangle inequality, and upper bound, and therefore is a distance metric in the space of ranked categorical features. It is also proven that the similarity measure presented in the study satisfies the axioms of boundedness, symmetry, maximum and minimum similarity, and is described by a decreasing function.
 Results. The developed approach has been implemented in an applied problem of determining the degree of similarity between objects described by ordered categorical features.
 Conclusions. In this study, mathematical tools were developed to determine similarity between structured data described by categorical attributes that can be ordered based on a specific priority in the form of a ranking system with preferences. Their properties were analyzed. Experimental studies have shown the convenience and “intuitive understanding” of the logic of data processing in solving practical problems. The proposed approach can provide the opportunity to conduct new meaningful research in data analysis. Prospects for further research lie in the experimental use of the proposed tools in practical tasks and in studying their effectiveness.

Clustering is a difficult and widely studied data mining task, with many varieties of clustering algorithms proposed in the literature. Nearly all algorithms use a similarity measure such as a distance metric (e.g., Euclidean distance) to decide which instances to assign to the same cluster. These similarity measures are generally predefined and cannot be easily tailored to the properties of a particular dataset, which leads to limitations in the quality and the interpretability of the clusters produced. In this article, we propose a new approach to automatically evolving similarity functions for a given clustering algorithm by using genetic programming. We introduce a new genetic programming-based method which automatically selects a small subset of features (feature selection) and then combines them using a variety of functions (feature construction) to produce dynamic and flexible similarity functions that are specifically designed for a given dataset. We demonstrate how the evolved similarity functions can be used to perform clustering using a graph-based representation. The results of a variety of experiments across a range of large, high-dimensional datasets show that the proposed approach can achieve higher and more consistent performance than the benchmark methods. We further extend the proposed approach to automatically produce multiple complementary similarity functions by using a multi-tree approach, which gives further performance improvements. We also analyse the interpretability and structure of the automatically evolved similarity functions to provide insight into how and why they are superior to standard distance metrics.

In real tasks, usually a good classification performance can only be obtained when a good distance metric is obtained; therefore, distance metric learning has attracted significant attention in the past few years. Typical studies of distance metric learning evaluate how to construct an appropriate distance metric that is able to separate training data points from different classes or satisfy a set of constraints (e.g., must-links and/or cannot-links). It is noteworthy that this task becomes challenging when there are only limited labeled training data points and no constraints are given explicitly. Moreover, most existing approaches aim to construct a global distance metric that is applicable to all data points. However, different data points may have different properties and may require different distance metrics. We notice that data points in real tasks are often connected by physical links (e.g., people are linked with each other in social networks; personal webpages are often connected to other webpages, including nonpersonal webpages), but the linkage information has not been exploited in distance metric learning. In this article, we develop a pairwised specific distance (PSD) approach that exploits the structures of physical linkages and in particular captures the key observations that nonmetric and clique linkages imply the appearance of different or unique semantics, respectively. It is noteworthy that, rather than generating a global distance, PSD generates different distances for different pairs of data points; this property is desired in applications involving complicated data semantics. We mainly present PSD for multi-class learning and further extend it to multi-label learning. Experimental results validate the effectiveness of PSD, especially in the scenarios in which there are very limited labeled training data points and no explicit constraints are given.