Nirattaya Khamsemanan

Human identification using skeletal gait and silhouette data extracted by Microsoft Kinect

Since the war on terrorists was declared, human identification area of research has gain its popularity throughout the world. Gait, a biométrie information obtained by ones walk, is used to identify a human widely because it can be done unobtrusively. Moreover, it is nearly impossible to alter gait features continuously. In this study, we propose a technique to identify a human using gait data extracted by Microsoft Kinect. We construct a distance function between two walking sequences using combinations of skeletal static features, skeletal kinematic features from movements and silhouette feature (mass vector). The proposed distance function is then used in the classification process along with fc-nearest neighbor technique. Our technique yields accuracy of 92.56% which outperforms those techniques proposed by Hong et. al., Cheewakidakarn et. al., Saitong-in et al., Preis et al., Milovanovic et al. and Boulgouris et al. Furthermore, we discover that skeletal kinematic features reveal the unique characteristic of human subjects better than skeletal static and silhouette features.

A novel prediction approach for antimalarial activities of Trimethoprim, Pyrimethamine, and Cycloguanil analogues using extremely randomized trees

Malaria is still one of the most serious diseases in tropical regions. This is due in part to the high resistance against available drugs for the inhibition of parasites, Plasmodium, the cause of the disease. New potent compounds with high clinical utility are urgently needed. In this work, we created a novel model using a regression tree to study structure-activity relationships and predict the inhibition constant, Ki of three different antimalarial analogues (Trimethoprim, Pyrimethamine, and Cycloguanil) based on their molecular descriptors. To the best of our knowledge, this work is the first attempt to study the structure-activity relationships of all three analogues combined. The most relevant descriptors and appropriate parameters of the regression tree are harvested using extremely randomized trees. These descriptors are water accessible surface area, Log of the aqueous solubility, total hydrophobic van der Waals surface area, and molecular refractivity. Out of all possible combinations of these selected parameters and descriptors, the tree with the strongest coefficient of determination is selected to be our prediction model. Predicted Ki values from the proposed model show a strong coefficient of determination, R2=0.996, to experimental Ki values. From the structure of the regression tree, compounds with high accessible surface area of all hydrophobic atoms (ASA_H) and low aqueous solubility of inhibitors (Log S) generally possess low Ki values. Our prediction model can also be utilized as a screening test for new antimalarial drug compounds which may reduce the time and expenses for new drug development. New compounds with high predicted Ki should be excluded from further drug development. It is also our inference that a threshold of ASA_H greater than 575.80 and Log S less than or equal to -4.36 is a sufficient condition for a new compound to possess a low Ki.

Fixed Points of Maps of a Nonaspherical Wedge

Let be a finite polyhedron that has the homotopy type of the wedge of the projective plane and the circle. With the aid of techniques from combinatorial group theory, we obtain formulas for the Nielsen numbers of the selfmaps of .

A prediction approach for anti-HIV activity of HEPT compounds using random forest technique

The human immunodeficiency virus type 1 (HIV-1) is one of the deadliest viruses that affect public health worldwide. Joint United Nations Programme on HIV/AIDS (UNAIDS) and World Health Organization estimate that there are more than 15 million people infected with this HIV-1 around the world. The cure of HIV-1 and a better understanding of effective drugs are urgently needed. In this work, we study the structure–activity relationship and predict the potency of HIV-1 drug compounds. We employ the random forest technique to select relevant molecular descriptors of 132 1-[(2-hydroxyethoxy)methyl]-6-(phenylthio)thymine (HEPT) compounds toward the inhibition of HIV-1 reverse transcriptase (RT). The best model yields 5 relevant descriptors with a coefficient of determination (R2) of 0.83. Our prediction model suggests that a potent HEPT compound must be a lipophilic molecule with a high value of fractional hydrophobic van der Waals surface areas.Graphical abstract

On the Black-box Use of Somewhat Homomorphic Encryption in NonInteractive Two-Party Protocols

In this work, we develop a methodology for determining the communication required to implement various two-party functionalities noninteractively. In the particular setting on which we focus, the protocols are based upon somewhat homomorphic encryption, and furthermore, they treat the homomorphic properties as a black box. In this setting, we develop lower bounds which give a smooth trade-off between the communication complexity and the “expressiveness” of the cryptosystem---the latter being measured in terms of the depth of the arithmetic circuits that can be evaluated on ciphertext. Given the current state of the art in homomorphic encryption, this trade-off may also be viewed as one between communication and computation, since at present, more expressive cryptosystems are markedly less efficient. We then apply this methodology to place lower bounds on a number of cryptographic protocols including private information retrieval writing and private keyword search. Our work provides a useful “litmus test” ...

A rule-based approach for improving Kinect Skeletal Tracking system with an application on standard Muay Thai maneuvers

The release of Xbox 360 Kinect has made a triumph impression of computer vision technology and open a new world of gaming experience. Behind this leading success is an intelligent recognition system. The fundamental process is the transformation of depth images into skeleton data. Though this system is not without its flaws. Kinect Skeletal Tracking system fails when it comes to rapidly moving sequences of postures, when parts of body are out of the line of sight of the camera or when the postures are not in natural human forms. Such movements can be found in standard Muay Thai maneuvers. In this paper, we propose an algorithm to improve the rate of skeleton recognition in such scenarios. With the Kinect recognition system alone, the accuracy of skeleton recognitions of all 24 standard Muay Thai maneuvers is 51%. After applying our proposed algorithm, we are able to bring the accuracy up 26%. On average, our proposed algorithm along with the Kinect recognition system yields accuracy of 77% on average standard Muay Thai maneuvers.

A fixed point theorem for smooth extension maps

AbstractLet X be a compact smooth n-manifold, with or without boundary, and let A be an (n−1)-dimensional smooth submanifold of the interior of X. Let ϕ:A→A be a smooth map and f:(X,A)→(X,A) be a smooth map whose restriction to A is ϕ. If p∈A is an isolated fixed point of f that is a transversal fixed point of ϕ, that is, the linear transformation dϕp−IA:TpA→TpA is nonsingular, then the fixed point index of f at p satisfies the inequality |i(X,f,p)|≤1. It follows that if ϕ has k fixed points, all transverse, and the Lefschetz number L(f)>k, then there is at least one fixed point of f in X∖A. Examples demonstrate that these results do not hold if the maps are not smooth. MSC:55M20, 54C20.

Interior fixed points of unit-sphere-preserving Euclidean maps

Schirmer proved that there is a class of smooth self-maps of the unit sphere in Euclidean n-space with the property that any smooth self-map of the unit ball that extends a map of that class must have at least one fixed point in the interior of the ball. We generalize Schirmer’s result by proving that a smooth self-map of Euclidean n-space that extends a self-map of the unit sphere of that class must have at least one fixed point in the interior of the unit ball.MSC:55M20, 54C20.

Human Identification From Freestyle Walks Using Posture-Based Gait Feature

With the increase of terrorist threats around the world, human identification research has become a sought after area of research. Unlike standard biometric recognition techniques, gait recognition is a non-intrusive technique. Both data collection and classification processes can be done without a subject’s cooperation. In this paper, we propose a new model-based gait recognition technique called postured-based gait recognition. It consists of two elements: posture-based features and posture-based classification. Posture-based features are composed of displacements of all joints between current and adjacent frames and center-of-body (CoB) relative coordinates of all joints, where the coordinates of each joint come from its relative position to four joints: hip-center, hip-left, hip-right, and spine joints, from the front forward. The CoB relative coordinate system is a critical part to handle the different observation angle issue. In posture-based classification, postured-based gait features of all frames are considered. The dominant subject becomes a classification result. The postured-based gait recognition technique outperforms the existing techniques in both fixed direction and freestyle walk scenarios, where turning around and changing directions are involved. This suggests that a set of postures and quick movements are sufficient to identify a person. The proposed technique also performs well under the gallery-size test and the cumulative match characteristic test, which implies that the postured-based gait recognition technique is not gallery-size sensitive and is a good potential tool for forensic and surveillance use.

Gait Recognition from Freestyle Walks Using Relative Coordinates and Random Subsequence-Based Sum-Rule Classification

Because of the increase in terrorist attacks, human identification research has become more popular. Gait recognition is one of biometric recognition techniques that can be accurately used to identify a person, since it is hard to alter the style of movement continuously and permanently. In this work, we propose a new gait recognition technique for freestyle walks using Microsoft Kinect. Since our technique supports freestyle walks, it can be employed to identify a person from a distance. This technique can also be used to collect gait information non-invasively and without a persons awareness. Our proposed technique introduce two new concepts to cope with challenges that come with freestyle walking. First, relative coordinate concept is created to handle the non-fixed observation angle issue. Second, the random subsequence-based sum-rule classification is introduced to handle non-fixed length walks issue. Our proposed technique takes different local characteristics of a walk into consideration. It significantly outperforms baseline technique (Dynamic Time Warping with k-NN with accuracy rate of 92.22%. The results also suggests that a human recognition is better done from observing many small movements and postures than observing one large sequence of walk.