• ETRI Journal
• /
• v.39 no.5
• /
• pp.643-651
• /
• 2017

Human age estimation is one of the key factors in the field of Human-Robot Interaction/Human-Computer Interaction (HRI/HCI). Owing to the development of deep-learning technologies, age recognition has recently been attempted. In general, however, deep learning techniques require a large-scale database, and for age learning with variations, a conventional database is insufficient. For this reason, we propose an age estimation method using artificially generated data. Image data are artificially generated through 3D information, thus solving the problem of shortage of training data, and helping with the training of the deep-learning technique. Augmentation using 3D has advantages over 2D because it creates new images with more information. We use a deep architecture as a pre-trained model, and improve the estimation capacity using artificially augmented training images. The deep architecture can outperform traditional estimation methods, and the improved method showed increased reliability. We have achieved state-of-the-art performance using the proposed method in the Morph-II dataset and have proven that the proposed method can be used effectively using the Adience dataset.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.6 no.2
• /
• pp.183-191
• /
• 2010

Decision problem called an optimal release policies, after testing a software system in development phase and transfer it to the user, is studied. The applied model of release time exploited infinite non-homogeneous Poisson process. This infinite non-homogeneous Poisson process is a model which reflects the possibility of introducing new faults when correcting or modifying the software. The failure life-cycle distribution used mixture which has various intensity, if the system is complicated. Thus, software release policies which minimize a total average software cost of development and maintenance under the constraint of satisfying a software reliability requirement becomes an optimal release policies. In a numerical example, after trend test applied and estimated the parameters using maximum likelihood estimation of inter-failure time data, estimated software optimal release time.

• Asia pacific journal of information systems
• /
• v.30 no.3
• /
• pp.457-496
• /
• 2020

This study aims to extensively analyze the performance of various Machine Learning (ML) techniques for predicting version to version change-proneness of source code Java files. 17 object-oriented metrics have been utilized in this work for predicting change-prone files using 31 ML techniques and the framework proposed has been implemented on various consecutive releases of two Java-based software projects available as plug-ins. 10-fold and inter-release validation methods have been employed to validate the models and statistical tests provide supplementary information regarding the reliability and significance of the results. The results of experiments conducted in this article indicate that the ML techniques perform differently under the different validation settings. The results also confirm the proficiency of the selected ML techniques in lieu of developing change-proneness prediction models which could aid the software engineers in the initial stages of software development for classifying change-prone Java files of a software, in turn aiding in the trend estimation of change-proneness over future versions.

• Convergence Security Journal
• /
• v.6 no.3
• /
• pp.145-153
• /
• 2006

Infinite failure NHPP models for a record value satisfies mode proposed in the literature exhibit either monotonic increasing or monotonic decreasing failure occurrence rates per fault. In this paper, propose comparative study of software reliability model using Erlang distribution, Rayleigh and Gumbel distribution. Equations to estimate the parameters using maximum likelihood estimation of infinite failure NHPP model based on failure data collected in the form of interfailure times are developed. For the sake of proposing distribution, we used to the special pattern. Analysis of failure data set using arithmetic and Laplace trend tests, goodness-of-fit test, bias tests is presented.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.27 no.10C
• /
• pp.887-901
• /
• 2002

Software management model divided into the software project model and design estimation model, software matrices model, reliability growth model, process improvement model(or process maturity model) etc. Among these software management models, software complexity model make an estimated of the product software. For a practice of software managed, need to guideline of the static analysis of software. Especially, Software complexity model introduced for the estimation of software quantity and program complexity. In case of measurement the software matrices, its need for us to analysis of software quality and products. On the other hand, we known that complexity program include many defects and consuming of source cost. So, we apply to complexity model using of the program complexity, control structure and volume matrices, interface metrics, process complexity metrics method. In this paper, we represent that the analysis of fault data detected during the system test. Also, we analysis of program control structure and interface, volume matrices in various aspect of switching software. Others, their results utilized similar of project and system development.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.9 no.6
• /
• pp.566-572
• /
• 2016

Software Reliability implemented in software development is one of the most important issues. In finite failure NHPP software reliability models for software failure analysis, the hazard function that means a failure rate may have constant independently for failure time, non-increasing or non-decreasing pattern. In this study, software development cost analysis considering the variable shape parameter of Erlang distribution as the failure life distribution in the software product testing process was studied. The software failure model was applied finite failure Non-Homogeneous Poisson Procedure and the parameters approximation using maximum likelihood estimation was accompanied. Thus, this paper was presented comparative analysis by applying a software failure time data to the software, considering the shape parameter of Erlang distribution for development cost model analysis. When compared to the cost curve in accordance with the shape parameter, the model of smaller shape can be seen that the optimal software release time delay and more cost. Through this study, it is thought that it can serve as a preliminary information which can basically help the software developers to search for development cost according to software shape parameters.

• The Transactions of the Korea Information Processing Society
• /
• v.5 no.9
• /
• pp.2345-2352
• /
• 1998

The hyper-geometric distribution software reliability growth model was recently developed and successfully applied Due to mathematical difficultv of the maximum likclihmd method, the least squares method has hem suggested for parameter estimation by the previous studies. We first summarize and compare the minimization criteria adopted by the previous studies. It is theo shown that the weighted least squares method is more appropriate hecause of the nonhomogeneous variability of the number of newly detected faults. The adequacy of the weighted least squares method is illustrated by two numerical examples. Finally, we propose a new method fur predicting the number of faults newly discovered by next test instances. The new prediction method can be used for determining the time to stop testing.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.33 no.2
• /
• pp.63-71
• /
• 2010

The existing reliability evaluation models which have already developed by the corporations are so various because of using Maximum Likelihood Method. The existing models are very complicated owing to using system designing methods. Therefore, it is very difficult to utilize the existing models in business fields of many corporations. The purposes of this paper are as follows: The first purpose is to study the simple estimated Parameter to be easily utilized in the business fields of the corporations. The second purpose is to testify the simplification of the developed Parameter of estimated method by comparing the developed reliability evaluation model with the existing reliability evaluation models which are used in the business fields of the corporations.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 2001.10a
• /
• pp.425-431
• /
• 2001

In order to take account of the statistical properties of probability variables used in the structural analysis, the conventional approach using the safety factor based on past experience usually estimated the safety of a structure. The real structures could only be analyzed with the error in estimation of loads, material characters and the dimensions of the members. But the errors should be considered systematically in the structural analysis. Structural safety could not precisely be appraised by the traditional structural design concept. Recently, new approach based on the probability concept has been applied to the assessment of structural safety using the reliability concept. In this study, safety of structures will estimated by the reliability analysis with commercial structural software that has the tools of nonlinear elastic-plastic 3-D analysis. Experimental test result is compared to results from this research and Coan/sup 1)/ In this paper, AFOSM(Advanced First-Order Second Moment method) is applied with von Mises, Tresca and Mohr-Coulomb failure criterions. The reliability index β and probability of failure P/sub f/ can be obtained by following this practical procedure as judgement a safety of structures and necessity of reinforcing.

• Journal of Korean Society for Quality Management
• /
• v.37 no.4
• /
• pp.16-22
• /
• 2009

The inflection S-shaped software reliability growth model (SRGM) proposed by Ohba(1984) is one of the most commonly used models and has been discussed by many authors. The main purpose of this paper is to estimate the parameters of Ohba's SRGM within the Bayesian framework by applying the Markov chain Monte Carlo techniques. While the maximum likelihood estimates for these parameters are well known, the Bayesian method for the inflection S-shaped SRGM have not been discussed in the literature. The proposed methods can be quite flexible depending on the choice of prior distributions for the parameters of interests. We also compare the Bayesian methods with the maximum likelihood method numerically based on the real data.