• MALSORI
• /
• no.51
• /
• pp.71-84
• /
• 2004

In this paper, we propose the automatic speech database verification method(or called automatic verification) based on confidence measure for a large speech database. This method verifies the consistency between given transcription and speech using the confidence measure. The automatic verification process consists of two stages : the word-level likelihood computation stage and multi-level likelihood ratio computation stage. In the word-level likelihood computation stage, we calculate the word-level likelihood using the viterbi decoding algorithm and make the segment information. In the multi-level likelihood ratio computation stage, we calculate the word-level and the phone-level likelihood ratio based on confidence measure with anti-phone model. By automatic verification, we have achieved about 61% error reduction. And also we can reduce the verification time from 1 month in manual to 1-2 days in automatic.

• Proceedings of the KSR Conference
• /
• 2010.06a
• /
• pp.1821-1828
• /
• 2010

In this study we suggest a method of optimization of verification hierarchy structure and system requirements management by using a verification matrix with traceability consideration. Verification items were gathered in the process of CDR(Critical Design Review), and analyzed with respect to requirements traceability structure. Missed or overlapped items were adjusted, and cross-correlated items between sub-systems were clustered and rearranged in order to structurize verification requrements (VRs). Those VRs are to be used as a guideline for the test and evaluation planning, development of test items and procedure, and system requirements management throughout the system integration stages.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.3C
• /
• pp.257-267
• /
• 2009

This paper presents a training method for neural networks and the employment of MSE (mean scare error) values as the basis of a decision regarding the identity claim of a speaker in a recurrent neural networks based speaker verification system. Recurrent neural networks (RNNs) are employed to capture temporally dynamic characteristics of speech signal. In the process of supervised learning for RNNs, target outputs are automatically generated and the generated target outputs are made to represent the temporal variation of input speech sounds. To increase the capability of discriminating between the true speaker and an impostor, a discriminative training method for RNNs is presented. This paper shows the use and the effectiveness of the MSE value, which is obtained from the Euclidean distance between the target outputs and the outputs of networks for test speech sounds of a speaker, as the basis of speaker verification. In terms of equal error rates, results of experiments, which have been performed using the Korean speech database, show that the proposed speaker verification system exhibits better performance than a conventional hidden Markov model based speaker verification system.

• Journal of the Korean Society of Systems Engineering
• /
• v.14 no.2
• /
• pp.16-23
• /
• 2018

The intended system's function and performance can be assured through implementing the development process, the verification compliance against corresponding requirements, in accordance with the fundamental principle from the Systems Engineering. For the effective verification implementation, related core metadata should be selected and managed throughout the development life cycle. And these have to be included in the configuration document such as specification so that taking them as development baselines each phases if necessary. In this paper, associated case study results are introduced to establish the Requirements Verification Matrix (RVM) for the verification management on the space launch vehicle development program.

• IE interfaces
• /
• v.25 no.1
• /
• pp.79-86
• /
• 2012

Traditional business process management systems provide verification tools of process models to deploy and automate the models. However, there are not so many studies on how to maintain systematically collaborative process models such as supply chain processes when companies are willing to change and update the collaborative process models. In this paper, we analyze change patterns of collaborative processes and declare 19 change patterns. In addition, we apply the change patterns to the process interoperability patterns in order to identify the change problems in case of independent process changes of collaborative processes. As a result, we devise an independency checking algorithm of process changes in collaborative processes.

• Journal of the Korean Society of Industry Convergence
• /
• v.22 no.6
• /
• pp.785-795
• /
• 2019

This paper is a study on the modeling of the quantity estimation model for offshore plant Material handling equipment in FEED(Front End Engineering Design) verification stage using system engineering approach which is an engineering design methods. The relevant engineering execution procedure is not systemized although the operation method and Material handling equipment selection with weight and space constraints is a key part of the FEED. Using the system engineering process, the stakeholder requirements analysis process, the system requirements analysis, and the final system architecture design were sequentially performed, and the process developed through the functional development diagram and Requirement traceability matrix (RTM) was verified. In addition, based on the established process, we propose a Material handling quantity estimation model and Quantity calculation verification Table that can be applied at the FEED verification stage and we verify the applicability through case studies.

• Korean Journal of Computational Design and Engineering
• /
• v.5 no.4
• /
• pp.359-372
• /
• 2000

Current CAD systems are good enough to be used as a tool to manipulate three-dimensional shapes. This is a very important capability to be owned by a design tool because a major portion of designers'activities is spent on the shape manipulation in the design detailing process. However, the whole design process involves a lot more than the, shape manipulation. Currently, these remaining tasks, mostly logical reasoning process for the function realization together with structure decomposition in the top-down manner, are processed in the designer's brain. To support the top-down functional design process of a mechanical product, a system integrating the functional, structural and geometrical aspects of a product design in a unified environment is presented. Using this system, a designer can perform function decomposition, structure decomposition, and geometry detailing, and function verification activities in parallel and the whole design process it modeled resultantly. Once the whole design process is modeled, any redesign task can be automatically performed with the verification of the desired functions.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.6 no.3
• /
• pp.247-254
• /
• 2006

Biometrics are getting more and more attention in recent years for security and other concerns. So far, only fingerprint recognition has seen limited success for on-line security check, since other biometrics verification and identification systems require more complicated and expensive acquisition interfaces and recognition processes. Hand-Geometry can be used for biometric verification and identification because of its acquisition convenience and good performance for verification and identification performance. It could also be a good candidate for online checks. Therefore, this paper proposes a Hand-Geometry recognition system based on geometrical features of hand. From anatomical point of view, human hand can be characterized by its length, width, thickness, geometrical composition, shapes of the palm, and shape and geometry of the fingers. This paper proposes thirty relevant features for a Hand-Geometry recognition system. This system presents verification results based on hand measurements of 20 individuals. The verification process has been tested on a size of $320{\times}240$ image, and result of the verification process have hit rate of 95% and FAR of 0.020.

• The Korean Journal of Archival Studies
• /
• no.77
• /
• pp.5-36
• /
• 2023

In the Republic of Korea, which has adopted a strong president-centered system as a political system, comprehensive personnel verification of work ability, reputation, and misconduct for positions where the president has personnel rights is not only fundamental to the administration of the country, but is also very important for maintaining the democratic system. As the records containing the personnel verification process produced during this process also has a wide range of data on important people in the country, they themselves are very important national-level information, and there is a great possibility that they will be used as useful materials for studying current national management in future generations. Therefore, it is assumed that after the enactment of the "Act on the Management of Presidential Records" in 2007, Senior Secretary for Civil Affairs of the Office of the President, which were responsible for the personnel verification, designated a wide range of personnel verification records as presidential designated records and transferred them to the Presidential Archives. However, with the launch of the 20th government, the agency responsible for personnel verification changed from Senior Secretary for Civil Affairs of the Office of the President to the "Personnel Information Management Team" of the Ministry of Justice from June 2022, a major change occurred in the personnel verification process that had been maintained and the preservation of related records. This paper examined what personnel verification records are and how they have been preserved and managed, and proposed what issues have arisen at the record management level due to this change in the agency in charge of personnel verification, and how those issues should be resolved on a practical and institutional level.

• Journal of Institute of Control, Robotics and Systems
• /
• v.19 no.12
• /
• pp.1099-1104
• /
• 2013

This paper presents a learning-based visual inspection method that addresses the need for an improved adaptability of a visual inspection system for parts verification in panorama sunroof assembly lines. It is essential to ensure that the many parts required (bolts and nuts, etc.) are properly installed in the PLC sunroof manufacturing process. Instead of human inspectors, a visual inspection system can automatically perform parts verification tasks to assure that parts are properly installed while rejecting any that are improperly assembled. The proposed visual inspection method is able to adapt to changing inspection tasks and environmental conditions through an efficient learning process. The proposed system consists of two major modules: learning mode and test mode. The SVM (Support Vector Machine) learning algorithm is employed to implement part learning and verification. The proposed method is very robust for changing environmental conditions, and various experimental results show the effectiveness of the proposed method.