Journal of Intelligent & Fuzzy Systems - Volume 36, issue 5

Authors: Tiwari, Anoop Kumar | Shreevastava, Shivam | Subbiah, Karthikeyan | Som, T.

Article Type: Research Article

Abstract:  Due to the development of modern internet-based technology, the electronically stored information is growing exponentially with time. It is highly challenging to select relevant and non-redundant features of the real-valued high dimensional datasets. Feature selection, a preprocessing technique, refers to the process of reducing the dimension of the input data in order to extract the most meaningful features for processing and analysis. One of the numerous useful applications of rough set theory is the attribute or feature selection, but it has certain limitations as it cannot be applied on real-valued data sets directly because rough set based feature selection can …handle discrete data only. In order to deal with real-valued data sets, discretization method is applied to convert dataset from real-valued to discrete, which usually leads to information loss. Fuzzy rough set theory is profitably applied to address this problem and retain the semantics of real-valued datasets. However, intuitionistic fuzzy set can deal with uncertainty in a much better way when compared to fuzzy set theory as it considers membership, non-membership and hesitancy degree of an object simultaneously. In this paper, an intuitionistic fuzzy rough set model is established by combining intuitionistic fuzzy set and rough set. Furthermore, we propose a novel approach of feature selection derived from this model. Moreover, we develop an algorithm based on our proposed concept. Finally, our approach is applied to some benchmark data sets and compared with the existing fuzzy rough set based technique. The performed experiments show the superiority of our approach. Show more

Keywords: Rough set, fuzzy set, intuitionistic fuzzy set, degree of dependency, feature selection

DOI: 10.3233/JIFS-179043

Citation: Journal of Intelligent & Fuzzy Systems, vol. 36, no. 5, pp. 4969-4979, 2019

Authors: Pinilla-Buitrago, Laura Alejandra | Carrasco-Ochoa, Jesús Ariel | Martínez-Trinidad, José Francisco

Article Type: Research Article

Abstract: Hieroglyph retrieval has emerged as a tool to facilitate and support the cultural heritage preservation. For this task, hieroglyphs should be represented according its visual content. In the literature, the Bag of Visual Words (BoVW) model has been widely used for representing hieroglyphs with retrieval purposes. One crucial step in the BoVW model consists in replacing each local descriptor, obtained from a hieroglyph, by its nearest visual word in the vocabulary. However, it may result in similar local descriptors replaced by different visual words. Thus, the similarity of these local descriptors is lost. In this work, this problem is …addressed by replacing each local descriptor by its k -nearest visual words in the vocabulary, instead of just one visual word (the nearest). Considering this multiple replacement, we introduce a hieroglyph representation that takes into account the frequency of the visual words and the co-occurrence of visual word pairs. Our experiments show that our proposed hieroglyph representation allows obtaining better retrieval results than those obtained by using state of the art representations. Show more

Keywords: Hieroglyph representation, hieroglyph retrieval, k-nearest visual words, co-occurrences of visual words pairs.

DOI: 10.3233/JIFS-179044

Citation: Journal of Intelligent & Fuzzy Systems, vol. 36, no. 5, pp. 4981-4990, 2019

Authors: Martínez-Espinosa, J.C. | Cordova-Fraga, T. | Guzmán-Cabrera, R.

Article Type: Research Article

Abstract: In this work, we propose a practical approach to access and visualize relevant information on the spatial distribution on the anything sample about its biochemical composition. In order to carry out this analysis, we use a Raman spectroscopy technique to obtain spectral maps with specific spatial resolution (1 and 5 micrometers) over a selected region of the sample. Our study relies on the application of a Principal Component Analysis on the cross-correlations between the spectral blocks measured, within a certain spectral window of interest. The associated values of these principal components are used to build low-resolution images (with the same …spatial resolution of the Raman scan) in which the relevant information on the chemical composition is already encoded. Finally, the spatial resolution of the principal components images was numerically enhanced in the post-processing through standard linear interpolation algorithms. In this way, we can map and visualize, simultaneously, the spatial and spectral information. The results suggest that the Raman spectroscopy imaging is a powerful tool for determining the biochemistry of organic and inorganic samples based on spectral scanning and thus determine compounds concentrations of medical interest. The proposed methodology is rather general and it could be extended to other spectroscopic measurement techniques where the spatial mapping of the spectral information is needed. Show more

Keywords: Raman spectroscopy, principal components analysis, spectral maps, image reconstruction

DOI: 10.3233/JIFS-179045

Citation: Journal of Intelligent & Fuzzy Systems, vol. 36, no. 5, pp. 4991-4999, 2019

Authors: Céspedes-Hernández, David | González-Calleros, Juan Manuel

Article Type: Research Article

Abstract: Human communication has been studied from different approaches and resulting in contributions to several disciplines. From the computer sciences point of view, the findings made in the area have inspired the development of Natural User Interfaces (NUI), interaction mechanisms aimed at replicating the way in which people communicate, so the information exchange with computational systems happens in similar fashion. Gestural interfaces are a specific type of NUI focused on analyzing the relationship between body motion and semantic meanings. Although, from a technical perspective, proposals found in the literature had proven high efficiency and accuracy on gestural recognition, several authors had …reported lack of naturalness in the interaction with gesture-based applications, leading to the conclusion that NUIs are not usually as natural as they claim to be. Moreover, gestures are culture and language specific, which makes them ambiguous, incompletely specified, and difficult to match with semantic meaning when the context is unknown. In this paper, we propose a methodology for enabling the development of gesture-based applications, considering that accuracy and efficiency in recognition tasks must not be affected, and prioritizing the flexibility for allowing the use of gestures that are suitable for different user contexts through the exploration of user-defined gesture sets and Machine Learning techniques, and using a one-shot learning approach. Show more

Keywords: Gestural interaction, natural user interfaces, human-computer interaction, machine learning, user-defined gesture sets

DOI: 10.3233/JIFS-179046

Citation: Journal of Intelligent & Fuzzy Systems, vol. 36, no. 5, pp. 5001-5010, 2019

Authors: Camarillo-Abad, Hector M. | Sanchez, J. Alfredo | Starostenko, Oleg | Sandoval, Maria Gabriela

Article Type: Research Article

Abstract:  Work is needed to advance the current understanding of tactile interaction among humans through haptic technologies. We introduce a novel language that has been designed to guide users in leader-follower dances. This language is based on a nine-word vocabulary that corresponds to nine dance steps, following the metaphor of a leader-follower dance. Our work benefits from a haptic coding that is commonly used by couples when dancing, and explores the potential of wearable technology in this scenario. A wearable prototype consisting of four vibrotactile actuators was used to test the idea. Two user studies show a high recognition rate (more …than 90%) of the intended tactile vocabulary. This particular work highlights the feasibility of a haptic vocabulary to exchange full, understandable commands between users, the importance of dance as a case study, and the potential of using wearable technology to support haptic communications in scenarios similar to those in the real world, such as partner dancing. Current results show it is viable to successfully guide someone to follow dance through communication using a basic vibrotactile language. Show more

Keywords: Wearable technology, haptic language, leader-follower dancing, natural user interfaces

DOI: 10.3233/JIFS-179047

Citation: Journal of Intelligent & Fuzzy Systems, vol. 36, no. 5, pp. 5011-5022, 2019

Authors: Aparicio-Díaz, Ernesto | Cumplido, René | Pérez Gort, Maikel Lázaro | Feregrino-Uribe, Claudia

Article Type: Research Article

Abstract:  One of the drawbacks of the current revolution on media is that tampering with images and videos is an increasingly easy task which brings a situation where digital media cannot be trusted. Digital video forensics study the effects of attacks and tampering techniques on videos and has arisen as a solution to the problem of lack of trustworthiness on digital media. Copy-Move tampering is one of the most common attacks, with variants for delete and duplicate objects in videos, and has been studied in several video forensics works with different approaches. Despite that, there is not yet a simple method …to determine multiple variants of Copy-Move attacks. This work proposes a simple yet effective method do detect Copy-Move for both subregion and full-frame duplication. Show more

Keywords: Signal processing, video processing, video forgery, video forensics, tampering detection, digital forensics

DOI: 10.3233/JIFS-179048

Citation: Journal of Intelligent & Fuzzy Systems, vol. 36, no. 5, pp. 5023-5035, 2019

Authors: Starostenko, Oleg | Cruz-Perez, Claudia | Alarcon-Aquino, Vicente | Rosas-Romero, Roberto

Article Type: Research Article

Abstract: In human-computer interaction the automatic face sensing and recognition of facial expressions is still a challenging task of affective computing, psychology and biomedical applications. The main goal of this paper is to increment a recognition rate of approaches for unobtrusive face sensing and automatic interpretation of emotions. The proposed approach explores local scale invariant feature transform descriptors for extraction of face key points used for face detection, recognition and then for encoding facial deformations in terms of Ekman’s Facial Action Coding System (FACS). Real-time face tracking and recognition is provided by quadratic discriminant analysis and Bayesian approaches as classification tools. …Based on detected fiducial points, the accurate automatic recognizing six prototypical human facial expressions as well as detecting affective states in real-time scenes is provided by fuzzy inference system based on the proposed reasoning model. Carried out experiments demonstrate that Ekman’s FACS traditionally used in affective computing may be extended to interpretation of non-prototypical compound emotions using Plutchik psychological model of emotional responses. Conducted tests with faces from standard databases confirm that the proposed approaches for analysis of local image features provide robust, quite accurate, fast and low computational cost face sensing and facial expression interpretation. Show more

Keywords: Affective computing, facial expression recognition, local face feature descriptors, fuzzy inference engine

DOI: 10.3233/JIFS-179049

Citation: Journal of Intelligent & Fuzzy Systems, vol. 36, no. 5, pp. 5037-5049, 2019

Authors: Pérez-Espinosa, Humberto | Torres-García, Alejandro Antonio

Article Type: Research Article

Abstract:  The barking and other vocalizations of the domestic dog are an exciting source of information. Studies in the area of ethology have analyzed their function and the way humans and conspecifics perceive them. Without a doubt, better understanding the nature of barking can bring benefits both, to improve the welfare of dogs, and for humans who can build systems that take advantage of the information extracted from vocalizations for applications, such as, security, assistance, and entertainment. To develop automatic systems for the analysis of domestic dog vocalizations, we need to have acoustic characterization methods that allow capturing the most relevant …properties of barking and thereby improving the performance of automatic classifiers. In this paper, a comparison between several acoustic characterization techniques is made to determine their relevance in the classification of two aspects of the barking, which are the context in which they were generated and the identity of the dog that emitted the bark. We classified the tested acoustic features as qualitative and quantitative. The quantitative are derived from the processing of low-level acoustic descriptors and have been used most widely in audio analysis. The qualitative ones are a type of acoustic that capture aspects related to the perception of the melody of the vocalizations and had not been previously tested in this field of application. Show more

Keywords: Dog’s vocalizations, acoustic features, automatic audio classification

DOI: 10.3233/JIFS-179050

Citation: Journal of Intelligent & Fuzzy Systems, vol. 36, no. 5, pp. 5051-5061, 2019

Authors: Herrera-Alcántara, Oscar | González-Mendoza, Miguel | Navarro-Fuentes, Jaime | Cruz-Barriguete, Víctor A.

Article Type: Research Article

Abstract:  Inverse parameterizations of length 12 orthogonal wavelet filters are presented, which allow to determine parameter values from filter coefficients. Its applicability is shown in a case of study of image processing where the optimization of five parameters is required. The parameterization of length N filters involves N 2 - 1 parameters, and it is easier to optimize shorter filters once they explore a subset of the search space. Under this approach, the optimization of length 12 filters is accelerated based on a nested optimization of length 4, 6, 8, and 10 filters by …exporting the best solutions from shorter to larger filters via inverse parameterizations. Experimental results support the success of the nested optimization when exploring the search space. The conclusions are that the use of the inverse formulas accelerates the convergence and that parameterized filters provide better results as their length increases and achieve a better performance than standard filters. Show more

Keywords: Wavelets, filter parameterization, image processing

DOI: 10.3233/JIFS-179051

Citation: Journal of Intelligent & Fuzzy Systems, vol. 36, no. 5, pp. 5063-5071, 2019

Authors: Francisco-Valencia, Iván | Marcial-Romero, José Raymundo | Valdovinos-Rosas, Rosa María

Article Type: Research Article

Abstract:  In this paper, we present a comparative analysis of two selection policies in the General Game Playing (GGP) context: Upper Confidence Bound (UCB) and Upper Confidence Bound Tuned (UCB-Tuned). The aim of the analysis is to identify which policy has the best performance in terms of victories in the GGP domain, a measure used in most of literature with other policies. In order to carry out the comparison, two agents were programmed using the GGP-base framework and the Monte Carlo Tree Search (MCTS) method. The games Breakthrough, Knightthrough and Connect Four were used as experimental scenarios, not compared previously to …the best of our knowledge. The results show that UCB-Tuned is better when less than 100 simulations are used in MCTS; however, when 1000 simulations are used, both policies have similar performance. Show more

Keywords: General Game Playing, Upper Confidence Bound, Upper Confidence Bound Tuned, Policies

DOI: 10.3233/JIFS-179052

Citation: Journal of Intelligent & Fuzzy Systems, vol. 36, no. 5, pp. 5073-5079, 2019