International Journal of Pattern Recognition and Artificial Intelligence

The bag-of-features model is a distinctive and robust approach to detect human actions in videos. The discriminative power of this model relies heavily on the quantization of the video features into visual words. The quantization determines how well the visual words describe the human action. Random forests have proven to efficiently transform the features into distinctive visual words. A major disadvantage of the random forest is that it makes binary decisions on the feature values, and thus not taking into account uncertainties of the values. We propose a soft-assignment random forest, which is a generalization of the random forest, by substitution of the binary decisions inside the tree nodes by a sigmoid function. The slope of the sigmoid models the degree of uncertainty about a feature's value. The results demonstrate that the soft-assignment random forest improves significantly the action detection accuracy compared to the original random forest. The human actions that are hard to detect — because they involve interactions with or manipulations of some (typically small) item — are structurally improved. Most prominent improvements are reported for a person handing, throwing, dropping, hauling, taking, closing or opening some item. Improvements are achieved for the state-of-the-art on the IXMAS and UT-Interaction datasets by using the soft-assignment random forest.

Feature selection has attracted a huge amount of interest in both research and application communities of hyperspectral image (HSI) classification. Generally, supervised feature selection methods are superior to unsupervised ones without label information. However, in classification of HSI, the labeled samples are often difficult, expensive or time-consuming to obtain. In this paper, we proposed a novel semi-supervised feature selection method, called Locality and Globality Discriminant Feature (LGDF), for HSI classification. This method combines Fisher's criteria and Graph Laplacian, which makes full use of both labeled and unlabeled data points to discover both manifold and discriminant structure in HSI data. In the proposed method, an optimal subset of features is identified if at this subset neighbor points or points sharing the same label are close to each other, while non-neighbor points or points with different labels are far away from each other. Experimental results on Washington DC Mall and AVIRIS Indian Pines hyperspectral datasets demonstrate the effectiveness of the proposed method.

This paper deals with the problems of language detection and tracking in multilingual online short word-of-mouth (WoM) discussions. This problem is particularly unusual and difficult from a pattern recognition perspective because, in these discussions, the participants and content involve the opinions of users from all over the world. The nature of these discussions, consisting of multiple topics in different languages, presents us with a problem of finding training and classification strategies when the class-conditional distributions are nonstationary. The difficulties in solving the problem are many-fold. First of all, the analyst has no knowledge of when one language stops and when the next starts. Further, the features which one uses for any one language (for example, the n-grams) will not be valid to recognize another. Finally, and most importantly, in most real-life applications, such as in WoM, the fragments of text available before the switching, are so small that it renders any meaningful classification using traditional estimation methods almost futile. Earlier, the authors [B. J. Oommen and L. Rueda, Patt. Recogn.39(1) (2006) 328–341.] had recommended that for a variety of problems, the use of strong estimators (i.e. estimators that converge with probability 1) is sub-optimal. In this vein, we propose to solve the current problem using novel estimators that are pertinent for nonstationary environments. The classification results obtained for various data sets which involve as many as eight languages demonstrates that our proposed methodology is both powerful and efficient.

Unprecedented advances in Internet technologies with multimedia capabilities have enabled pornography and adult content to be widely and freely distributed as easy as a click of a mouse through various means such as YouTube, Facebook, and Tags. Protecting children from unnecessary exposure to adult content has, therefore, become a serious problem in the real world. In particular, the considerable perversion in pornography and the exposure of children and the society to such perversions leads to moral decay. Constructing an appropriate filter for pornographic images is a major concern in modern society; however, this area poses challenges. This study aims to shed light on a content-based technique that employs an anti-pornography machine and to encourage researchers to study this adult image filtering technique. In this study, we discuss models of skin detection and their advantages and disadvantages in real life. We also elaborate on the pornographic image classifier using a feature extraction process and its classification process, along with the possible difficulties it may present. This study also analyzes anti-pornography techniques based on skin detection and discusses their strengths and weaknesses.

In the last decade, new approaches focused on modeling uncertainty over complex relational data have been developed. In this paper, one of the most promising of such approaches, known as probabilistic relational model (PRM), has been investigated and extended in order to measure and include semantic relationships for addressing web page classification problems. Experimental results show the potential of the proposed method of capturing the "strength" of existing relationships (links) and the capacity of including this information into the probability model.

In this paper, a post processor for accuracy of character recognition of real-time online Gurmukhi script has been developed. Our analysis is based on dataset consisting of 184 samples of each 45 characters of Gurmukhi script collected from four different categories of writers. Based on this extensive study, we propose an efficient algorithm for online handwritten Gurmukhi character recognition that achieves promising recognition accuracy of 95.6% for single character stroke sequencing. Beside character recognition the contribution in this paper is summarized in two folds as (i) the proposed scheme resolves stroke sequencing, (ii) overwritten strokes are identified and resolved. Moreover, for every stroke, complexity of adding new stroke for Gurmukhi character formation has been computed to be O(n).

Steerability is a useful and important property of "kernel" functions. It enables certain complicated operations involving orientation manipulation on images to be executed with high efficiency. Thus, we focus our attention on the steerability of Hermite polynomials and their versions modulated by the Gaussian function with different powers, defined as the Hermite kernel. Certain special cases of such kernel, Hermite polynomials, Hermite functions and Gaussian derivatives are discussed in detail. Correspondingly, these cases demonstrate that the Hermite kernel is a powerful and effective tool for image processing. Furthermore, the steerability of the Hermite kernel is proved with the help of a property of Hermite polynomials revealing the rule concerning the product of two Hermite polynomials after coordination rotation. Consequently, any order of the Hermite kernel inherits steerability. Moreover, a couple sets of an explicit interpolation function and basis function can be directly obtained. We provide some examples to verify steerability of the Hermite kernel. Experimental results show the effectiveness of steerability and its potential applications in the fields of image processing and computer vision.

Face recognition has become more significant and relevant in recent years owing to it potential applications. Since the faces are highly dynamic and pose more issues and challenges to solve, researchers in the domain of pattern recognition, computer vision and artificial intelligence have proposed many solutions to reduce such difficulties so as to improve the robustness and recognition accuracy. As many approaches have been proposed, efforts are also put in to provide an extensive survey of the methods developed over the years. The objective of this paper is to provide a survey of face recognition papers that appeared in the literature over the past decade under all severe conditions that were not discussed in the previous survey and to categorize them into meaningful approaches, viz. appearance based, feature based and soft computing based. A comparative study of merits and demerits of these approaches have been presented.

In this paper, a biometric technique based on gesture recognition is proposed to improve the security of operations requiring authentication in mobile phones. Users are authenticated by making a gesture invented by them holding a mobile phone embedding an accelerometer on their hand. An analysis method based on sequence alignment is proposed and evaluated in different experiments. Firstly, a test of distinctiveness of gestures has been proposed obtaining an equal error rate (EER) of 4.98% with a database of 30 users and four repetitions. With the same database, a second experiment representing the unicity of accessing attempts has resulted in an EER value of 1.92%. Finally, a third experiment to evaluate the robustness of the technique has examined a database of 40 users with eight repetitions and real falsification attempts, performed by three impostors from the study of recordings of the carrying out of the original gestures, resulting in an EER of 2.5%.

Diabetes mellitus is a common chronic disease in recent years. According to the World Health Organization, the estimated number of diabetic patients will increase 56% in Asia from the year 2010 to 2025, where the number of anti-diabetic drugs that doctors are able to utilize also increase as the development of pharmaceutical drugs. In this paper, we present a recommendation system for anti-diabetic drugs selection based on fuzzy reasoning and ontology techniques, where fuzzy rules are used to represent knowledge to infer the usability of the classes of anti-diabetic drugs based on fuzzy reasoning techniques. We adopt the "Medical Guidelines for Clinical Practice for the Management of Diabetes Mellitus" provided by the American Association of Clinical Endocrinologists to build the ontology knowledge base. The experimental results show that the proposed anti-diabetic drugs recommendation system gets the same accuracy rate as the one of Chen et al.'s method (R. C. Chen, Y. H. Huang, C. T. Bau and S. M. Chen, Expert Syst. Appl.39(4) (2012) 3995–4006.) and it is better than Chen et al.'s method (R. C. Chen, Y. H. Huang, C. T. Bau and S. M. Chen, Expert Syst. Appl.39(4) (2012) 3995–4006.) due to the fact that it can deal with the semantic degrees of patients' tests and can provide different recommend levels of anti-diabetic drugs. It provides us with a useful way for anti-diabetic drugs selection based on fuzzy reasoning and ontology techniques.