• Journal of Korea Society of Industrial Information Systems
• /
• v.16 no.4
• /
• pp.9-18
• /
• 2011

In this paper, we propose a method which recognizes human activity using model-based gaze direction estimation in an indoor environment. The method consists of two steps. First, we detect a head region and estimate its gaze direction as prior information in the human activity recognition. We use color and shape information for the detection of head region and use Bayesian Network model representing relationships between a head and a face for the estimation of gaze direction. Second, we recognize event and scenario describing the human activity. We use change of human state for the event recognition and use a rule-based method with combination of events and some constraints. We define 4 types of scenarios related to the gaze direction. We show performance of the gaze direction estimation and human activity recognition with results of experiments.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.44 no.6
• /
• pp.16-23
• /
• 2007

Changeable biometrics has been developed as a solution to problem of enhancing security and privacy. The idea is to transform a biometric signal or feature into a new one for the purposes of enrollment and matching. In this paper, we propose a changeable biometric system that can be applied to appearance based face recognition system. In the first step when using feature extraction, ICA(Independent Component Analysis) coefficient vectors extracted from an input face image are replaced randomly using their mean and variation. The transformed vectors by replacement are scrambled randomly and a new transformed face coefficient vector (transformed template) is generated by combination of the two transformed vectors. When this transformed template is compromised, it is replaced with new random numbers and a new scrambling rule. Because e transformed template is generated by e addition of two vectors, e original ICA coefficients could not be easily recovered from the transformed coefficients.

• Speech Sciences
• /
• v.10 no.4
• /
• pp.201-219
• /
• 2003

Background: The perceptual assessment is generally performed by the voice specialist. The objective evaluation is performed in a voice laboratory. Research in voice laboratories has generated a variety of different objective tests and parameters. The perceptual evaluation is one of the most controversial topics in voice research. Review of literature reveals a wide variety of rating scales and reliability data fluctuating from study to study. Unfortunately, there is no widely accepted valid method for classifying voice disorders and assessing outcome after voice treatment. Objectives: The goals of this research were to identify important objective acoustic parameters of vocal quality, and to establish an objective and quantitative correlate of the perceived vocal quality. Materials and Methods : We evaluated the voice analyzed data from 122 dysphonic patients and 20 normal volunteers. A computerized speech lab. 4300B(CSL) was used to carry out the analysis of each voice sample. Results: Three dysphonia severity indices(DSI) were created using discriminant analysis. DSI is based on the weighted combination of the following selected set of acoustic parameters: absolute jitter(Jita in us), smoothed pitch period perturbation (sPPQ in %), amplitude perturbation quotient(APQ in %), soft phonation index(SPI), average fundamental frequency(Fo in Hz), lowest fundamental frequency(Flo in Hz), and smoothed amplitude perturbation quotient(sAPQ in %). The DSI, being the discriminating rule calculated by the logistic regression, consists of three equation based on statistically significant acoustic parameters. Three DSI were created to reflects best the degree of hoarseness as expressed by G from the GRBAS scale. The more positive this DSI is for a patient, the worse the vocal quality. The more it is negative, the better it is. The effect of sex is included implicitly in the DSI-1 and DSI-2, so that a separate DSI-1 and DSI-2 for males and females need not be used. The DSI is objective because no perceptual input is required for its calculation. Conculsion : This research demonstrates that the voice function values calculated from three different multivariate objective dysphonia severity indices are significantly associated with subjective voice assessments. These multivariate objective dysphonia severity indices may be appropriate for use in clinical trials and outcomes research on treatment effectiveness for voice disorders.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2015.10a
• /
• pp.808-811
• /
• 2015

Cognitive radio has been proposed as a promising dynamic spectrum allocation paradigm. In cognitive radio, spectrum sensing is a fundamental procedure that enables secondary users (unlicensed) employing unused portion of spectrum of primary users (licensed) without causing harmful interference. However, the performance of single-user spectrum-sensing scheme was limited by fading, noise uncertainty shadowing and hidden node problem. Cooperative spectrum sensing was proposed to mitigate these problem. In this paper, we observe cooperative sensing scheme with energy detection using three adaptive thresholds for local decision, which can mitigate sensing failure problem and improve sensing performance at local node. In cooperative scheme we employed OR rules as decision combining at fusion center. We evaluate our scheme through computer simulation, and the results show that with OR combination rule our scheme can achieve best performance than other schemes.

• Structural Engineering and Mechanics
• /
• v.12 no.5
• /
• pp.507-526
• /
• 2001

In standard finite element algorithms, the local stability conditions are not accounted for in the formulation of the tangent stiffness matrix. As a result, the loss of the local stability is not adequately related to the onset of the global instability. The phenomenon typically arises with material-type localizations, such as shear bands and plastic hinges. This paper addresses the problem in the context of the planar, finite-strain, rate-independent, materially non-linear beam theory, although the proposed technology is in principle not limited to beam structures. A weak formulation of Reissner's finite-strain beam theory is first presented, where the pseudocurvature of the deformed axis is the only unknown function. We further derive the local stability conditions for the large deformation case, and suggest various possible combinations of the interpolation and numerical integration schemes that trigger the simultaneous loss of the local and global instabilities of a statically determined beam. For practical applications, we advice on a procedure that uses a special numerical integration rule, where interpolation nodes and integration points are equal in number, but not in locations, except for the point of the local instability, where the interpolation node and the integration point coalesce. Provided that the point of instability is an end-point of the beam-a condition often met in engineering practice-the procedure simplifies substantially; one of such algorithms uses the combination of the Lagrangian interpolation and Lobatto's integration. The present paper uses the Galerkin finite element discretization, but a conceptually similar technology could be extended to other discretization methods.

• Journal of KIISE
• /
• v.43 no.1
• /
• pp.80-86
• /
• 2016

A speech-act is a behavior intended by users in an utterance. Speech-act classification is important in a dialogue system. The machine learning and rule-based methods have mainly been used for speech-act classification. In this paper, we propose a speech-act classification method based on the combination of support vector machine (SVM) and transformation-based learning (TBL). The user's utterance is first classified by SVM that is preferentially applied to categories with a low utterance rate in training data. Next, when an utterance has negative scores throughout the whole of the categories, the utterance is applied to the correction phase by rules. The results from our method were higher performance over the baseline system long with error-reduction.

• Journal of KIISE
• /
• v.43 no.7
• /
• pp.781-785
• /
• 2016

Learning from human behaviors in the real world is essential for human-aware intelligent systems such as smart assistants and autonomous robots. Most of research focuses on correlations between sensory patterns and a label for each activity. However, human activity is a combination of several event contexts and is a narrative story in and of itself. We propose a novel approach of human activity prediction based on event cognition. Egocentric multi-sensor data are collected from an individual's daily life by using a wearable device and smartphone. Event contexts about location, scene and activities are then recognized, and finally the users" daily activities are predicted from a decision rule based on the event contexts. The proposed method has been evaluated on a wearable sensor data collected from the real world over 2 weeks by 2 people. Experimental results showed improved recognition accuracies when using the proposed method comparing to results directly using sensory features.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.9 no.10
• /
• pp.3965-3982
• /
• 2015

Event detection is one of the key issues in many wireless sensor network (WSN) applications. The uncertainties that are derived from the instability of sensor node, measurement noise and incomplete sampling would influence the performance of event detection to a large degree. Many of the present researches described the sensor readings with crisp values, which cannot adequately handle the uncertainties inhered in the imprecise sensor readings. In this paper, a fault-tolerant event detection algorithm is proposed based on Dempster-Shafer (D-S) theory (also called evidence theory). Instead of crisp values, all possible states of the event are represented by the Basic Probability Assignment (BPA) functions, with which the output of each sensor node are characterized as weighted evidences. The combination rule was subsequently applied on each sensor node to fuse the evidences gathered from the neighboring nodes to make the final decision on whether the event occurs. Simulation results show that even 20% nodes are faulty, the accuracy of the proposed algorithm is around 80% for event region detection. Moreover, 97% of the error readings have been corrected, and an improved detection capability at the boundary of the event region is gained by 75%. The proposed algorithm can enhance the detection accuracy of the event region even in high error-rate environment, which reflects good reliability and robustness. The proposed algorithm is also applicable to boundary detection as it performs well at the boundary of the event.

• Korean Journal of Community Nutrition
• /
• v.8 no.1
• /
• pp.62-70
• /
• 2003

This article aims to summarize the development and application of menu engineering technique, 'Menu Engineering Modified by Preference (MEMP)'. The site selected for this project was a foodservice operation in Yonsei University residence hall. Sales and food costs data were collected from the daily sales reports for 1 month, and the survey of food preference was conducted during May, 1999. Statistical data analysis was completed using the SAS/Win 6.12 for descriptive analysis. The calculation for menu analysis were carried out with MS 2000 Excel spreadsheet program. This MEMP technique developed had 6 category criteria and 2 dimensions of the contribution margin (CM) and the menu mix modified% (MMM%) . The MMM% was calculated by the sales volumes and also weighted by food preference. The CM and MMM% for each item were compared with a mean menu CM as well as a 70% rule. Four possible classifications by MEMP were fumed out as 'STAR', 'PLOWHORSE', 'PUZZLE', 'DOG'. 'STAR' items were the most popular and profitable items and required to maintain rigid specifications for quality. The decision actions for 'PLOWHORSE' menu items which were relatively popular, but yield a low menu average CM included combining a plowhorse item with lower cost products and reducing the frequency of serving or serving size. There was a need for 'PUZZLE' items to be changed in the menu combination, improve recipe, and promote menu. The last DOG' items were desired to be deleted. This study demonstrates that menu information can be interpreted more easily with MEMP. The use of MEMP is therefore an effective way to improve management decisions about menu of university residence hall foodservice.

• Journal of the Korean Society for Library and Information Science
• /
• v.49 no.3
• /
• pp.81-107
• /
• 2015

This study describes the Bottom-up method for implementation of an ontology system from the RDB. The STNet, a structured terminology glossary based on RDB, was served as a test bed for converting to RDF ontology, for generating the inference rules, and for evaluating the results of the semantic search. We have used protege editor of the ontology developing tool to design ontologies with test data. We also tested the designed ontology with the Inference Engine (Pellet) of protege editor. The generated reference rules were tested by TBox and SPARQL queries through STNet ontology. The results of test show that the generated reference rules were verified as true and STNet ontology were also evaluated to be useful for searching the complex combination of semantic relation.