Subject independent human action recognition using spatio-depth information and meta-cognitive RBF network

In this paper, we present a machine learning approach for subject independent human action recognition using depth camera, emphasizing the importance of depth in recognition of actions. The proposed approach uses the flow information of all 3 dimensions to classify an action. In our approach, we have obtained the 2-D optical flow and used it along with the depth image to obtain the depth flow (Z motion vectors). The obtained flow captures the dynamics of the actions in space time. Feature vectors are obtained by averaging the 3-D motion over a grid laid over the silhouette in a hierarchical fashion. These hierarchical fine to coarse windows capture the motion dynamics of the object at various scales. The extracted features are used to train a Meta-cognitive Radial Basis Function Network (McRBFN) that uses a Projection Based Learning (PBL) algorithm, referred to as PBL-McRBFN, henceforth. PBL-McRBFN begins with zero hidden neurons and builds the network based on the best human learning strategy, namely, self-regulated learning in a meta-cognitive environment. When a sample is used for learning, PBLMcRBFN uses the sample overlapping conditions, and a projection based learning algorithm to estimate the parameters of the network. The performance of PBL-McRBFN is compared to that of a Support Vector Machine (SVM) and Extreme Learning Machine (ELM) classifiers with representation of every person and action in the training and testing datasets. Performance study shows that PBL-McRBFN outperforms these classifiers in recognizing actions in 3-D. Further, a subject-independent study is conducted by leave-one-subject-out strategy and its generalization performance is tested. It is observed from the subject-independent study that McRBFN is capable of generalizing actions accurately. The performance of the proposed approach is benchmarked with Video Analytics Lab (VAL) dataset and Berkeley Multimodal Human Action Database (MHAD). (C) 2013 Elsevier Ltd. All rights reserved.

An Optical How Feature and McFIS Based Approach for 3-Dimensional Human Action Recognition

We propose to develop a 3-D optical flow features based human action recognition system. Optical flow based features are employed here since they can capture the apparent movement in object, by design. Moreover, they can represent information hierarchically from local pixel level to global object level. In this work, 3-D optical flow based features a re extracted by combining the 2-1) optical flow based features with the depth flow features obtained from depth camera. In order to develop an action recognition system, we employ a Meta-Cognitive Neuro-Fuzzy Inference System (McFIS). The m of McFIS is to find the decision boundary separating different classes based on their respective optical flow based features. McFIS consists of a neuro-fuzzy inference system (cognitive component) and a self-regulatory learning mechanism (meta-cognitive component). During the supervised learning, self-regulatory learning mechanism monitors the knowledge of the current sample with respect to the existing knowledge in the network and controls the learning by deciding on sample deletion, sample learning or sample reserve strategies. The performance of the proposed action recognition system was evaluated on a proprietary data set consisting of eight subjects. The performance evaluation with standard support vector machine classifier and extreme learning machine indicates improved performance of McFIS is recognizing actions based of 3-D optical flow based features.

Database Independent Human Emotion Recognition with Meta-Cognitive Neuro-Fuzzy Inference System

Facial emotions are the most expressive way to display emotions. Many algorithms have been proposed which employ a particular set of people (usually a database) to both train and test their model. This paper focuses on the challenging task of database independent emotion recognition, which is a generalized case of subject-independent emotion recognition. The emotion recognition system employed in this work is a Meta-Cognitive Neuro-Fuzzy Inference System (McFIS). McFIS has two components, a neuro-fuzzy inference system, which is the cognitive component and a self-regulatory learning mechanism, which is the meta-cognitive component. The meta-cognitive component, monitors the knowledge in the neuro-fuzzy inference system and decides on what-to-learn, when-to-learn and how-to-learn the training samples, efficiently. For each sample, the McFIS decides whether to delete the sample without being learnt, use it to add/prune or update the network parameter or reserve it for future use. This helps the network avoid over-training and as a result improve its generalization performance over untrained databases. In this study, we extract pixel based emotion features from well-known (Japanese Female Facial Expression) JAFFE and (Taiwanese Female Expression Image) TFEID database. Two sets of experiment are conducted. First, we study the individual performance of both databases on McFIS based on 5-fold cross validation study. Next, in order to study the generalization performance, McFIS trained on JAFFE database is tested on TFEID and vice-versa. The performance The performance comparison in both experiments against SVNI classifier gives promising results.

Hydrogen-bond-directed self-assembly of D-(+)-dibenzoyltartaric acid and 4-aminopyridine: optical nonlinearities and stoichiometry-dependent novel structural features

Attempts to prepare hydrogen-bond-directed nonlinear optical materials from a 1:1 molar mixture Of D-(+)-dibenzoyltartaric acid (DBT, I) and 4-aminopyridine (4-AP, II) resulted in two salts of different stoichiometry. One of them crystallizes in an unusual 1.5:1 (acid:base) monohydrate salt form III while the other one crystallizes as 1:1 (acid:base) salt IV. Crystal structures of both of the salts were determined from single-crystal X-ray diffraction data. The salt III crystallizes in a monoclinic space group C2 with a = 30.339(l), b = 7.881(2), c = 14.355(1) angstrom, beta = 97.48(1)degrees, V = 3403.1(9) angstrom3, Z = 4, R(w) = 0.058, R(w)= 0.058. The salt IV also crystallizes in a monoclinic space group P2(1) with a = 7.500(1), b = 14.968(2), c = 10.370(1) angstrom, beta = 102.67(1)degrees, V = 1135.9(2) angstrom3, Z = 2, R = 0.043, R(w) = 0.043. Interestingly, two DBT molecules with distinctly different conformation are present in the same crystal lattice of salt III. Extensive hydrogen-bonding interactions are found in both of the salts, and both of them show SHG intensity 1.4-1.6 times that of urea.

Shape and size directed self-selection in organic cage formation

The selective formation of a single isomer of a 3+2] self-assembled organic cage from a reaction mixture of an unsymmetrical aldehyde and a flexible amine is discussed. The experimental and theoretical findings suggest that in such a process, the geometric features of the aldehyde play a key role.

An innovative clustering technique for unsupervized learning in the context of remotely sensed earth resources data analysis

A new clustering technique, based on the concept of immediato neighbourhood, with a novel capability to self-learn the number of clusters expected in the unsupervized environment, has been developed. The method compares favourably with other clustering schemes based on distance measures, both in terms of conceptual innovations and computational economy. Test implementation of the scheme using C-l flight line training sample data in a simulated unsupervized mode has brought out the efficacy of the technique. The technique can easily be implemented as a front end to established pattern classification systems with supervized learning capabilities to derive unified learning systems capable of operating in both supervized and unsupervized environments. This makes the technique an attractive proposition in the context of remotely sensed earth resources data analysis wherein it is essential to have such a unified learning system capability.

A self-adaptive migration model genetic algorithm for data mining applications

Data mining involves nontrivial process of extracting knowledge or patterns from large databases. Genetic Algorithms are efficient and robust searching and optimization methods that are used in data mining. In this paper we propose a Self-Adaptive Migration Model GA (SAMGA), where parameters of population size, the number of points of crossover and mutation rate for each population are adaptively fixed. Further, the migration of individuals between populations is decided dynamically. This paper gives a mathematical schema analysis of the method stating and showing that the algorithm exploits previously discovered knowledge for a more focused and concentrated search of heuristically high yielding regions while simultaneously performing a highly explorative search on the other regions of the search space. The effective performance of the algorithm is then shown using standard testbed functions and a set of actual classification datamining problems. Michigan style of classifier was used to build the classifier and the system was tested with machine learning databases of Pima Indian Diabetes database, Wisconsin Breast Cancer database and few others. The performance of our algorithm is better than others.

A Dynamic Migration Model for Self-adaptive Genetic Algorithms

In this paper, we propose a self Adaptive Migration Model for Genetic Algorithms, where parameters of population size, the number of points of crossover and mutation rate for each population are fixed adaptively. Further, the migration of individuals between populations is decided dynamically. This paper gives a mathematical schema analysis of the method stating and showing that the algorithm exploits previously discovered knowledge for a more focused and concentrated search of heuristically high yielding regions while simultaneously performing a highly explorative search on the other regions of the search space. The effective performance of the algorithm is then shown using standard testbed functions, when compared with Island model GA(IGA) and Simple GA(SGA).

A tutorial survey of reinforcement learning

This paper gives a compact, self-contained tutorial survey of reinforcement learning, a tool that is increasingly finding application in the development of intelligent dynamic systems. Research on reinforcement learning during the past decade has led to the development of a variety of useful algorithms. This paper surveys the literature and presents the algorithms in a cohesive framework.

Reducing convergence times of self-propelled swarms via modified nearest neighbor rules

Vicsek et al. proposed a biologically inspired model of self-propelled particles, which is now commonly referred to as the Vicsek model. Recently, attention has been directed at modifying the Vicsek model so as to improve convergence properties. In this paper, we propose two modification of the Vicsek model which leads to significant improvements in convergence times. The modifications involve an additional term in the heading update rule which depends only on the current or the past states of the particle's neighbors. The variation in convergence properties as the parameters of these modified versions are changed are closely investigated. It is found that in both cases, there exists an optimal value of the parameter which reduces convergence times significantly and the system undergoes a phase transition as the value of the parameter is increased beyond this optimal value. (C) 2012 Elsevier B.V. All rights reserved.

On the ultrafast charge migration and subsequent charge directed reactivity in Cl center dot center dot center dot N halogen-bonded clusters following vertical ionization

In this article, we have presented ultrafast charge transfer dynamics through halogen bonds following vertical ionization of representative halogen bonded clusters. Subsequent hole directed reactivity of the radical cations of halogen bonded clusters is also discussed. Furthermore, we have examined effect of the halogen bond strength on the electron-electron correlation-and relaxation-driven charge migration in halogen bonded complexes. For this study, we have selected A-Cl (A represents F, OH, CN, NH2, CF3, and COOH substituents) molecules paired with NH3 (referred as ACl:NH3 complex): these complexes exhibit halogen bonds. To the best of our knowledge, this is the first report on purely electron correlation-and relaxation-driven ultrafast (attosecond) charge migration dynamics through halogen bonds. Both density functional theory and complete active space self-consistent field theory with 6-31+G(d, p) basis set are employed for this work. Upon vertical ionization of NCCl center dot center dot center dot NH3 complex, the hole is predicted to migrate from the NH3-end to the ClCN-end of the NCCl center dot center dot center dot NH3 complex in approximately 0.5 fs on the D-0 cationic surface. This hole migration leads to structural rearrangement of the halogen bonded complex, yielding hydrogen bonding interaction stronger than the halogen bonding interaction on the same cationic surface. Other halogen bonded complexes, such as H2NCl:NH3, F3CCl:NH3, and HOOCCl:NH3, exhibit similar charge migration following vertical ionization. On the contrary, FCl:NH3 and HOCl:NH3 complexes do not exhibit any charge migration following vertical ionization to the D-0 cation state, pointing to interesting halogen bond strength-dependent charge migration. (C) 2015 AIP Publishing LLC.