• Phonetics and Speech Sciences
• /
• v.15 no.3
• /
• pp.75-82
• /
• 2023

Automatic speech recognition (ASR) has been revolutionized with deep learning-based approaches, among which self-supervised learning methods have proven to be particularly effective. In this study, we aim to enhance the performance of OpenAI's Whisper model, a multilingual ASR system on the Korean language. Whisper was pretrained on a large corpus (around 680,000 hours) of web speech data and has demonstrated strong recognition performance for major languages. However, it faces challenges in recognizing languages such as Korean, which is not major language while training. We address this issue by fine-tuning the Whisper model with an additional dataset comprising about 1,000 hours of Korean speech. We also compare its performance against a Transformer model that was trained from scratch using the same dataset. Our results indicate that fine-tuning the Whisper model significantly improved its Korean speech recognition capabilities in terms of character error rate (CER). Specifically, the performance improved with increasing model size. However, the Whisper model's performance on English deteriorated post fine-tuning, emphasizing the need for further research to develop robust multilingual models. Our study demonstrates the potential of utilizing a fine-tuned Whisper model for Korean ASR applications. Future work will focus on multilingual recognition and optimization for real-time inference.

• Journal of KIISE:Databases
• /
• v.41 no.4
• /
• pp.209-216
• /
• 2014

Various characteristics of social network contents such as real-time, people relationship and big data can help to improve personalized recommender systems. Among them, 'people relationship' is a key factor of recommendation, so many personalized recommender systems utilizing it have been proposed. However, existing researches can not reflect personal tendency and are unable to provide precise recommendations in various domains, because they do not consider intimacy among people. In this paper, to solve these problems, we propose PReAmacy, a Personalized Recommendation Algorithm, considering intimacy among users and various characteristics of social network contents. Our experimental results indicate that not only the precision of PReAmacy is higher than that of existing algorithms, but intimacy is of great importance in PReAmacy.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.11 no.10
• /
• pp.3873-3879
• /
• 2010

Home service robot are not working in the fixed task such as industrial robot, because they are together with human in the same indoor space, but have to do in much more flexible and various environments. Most of them are developed on the base of the wheel-base mobile robot in the same method as a vehicle robot for factory automation. In these days, for holonomic system characteristics, omni-directional wheels are used in the mobile robot. A holonomicrobot, using omni-directional wheels, is capable of driving in any direction. But trajectory control for omni-directional mobile robot is not easy. Especially, azimuth control which sensor uncertainty problem is included is much more difficult. This paper develops trajectory controller of 3-wheels omni-directional mobile robot using fuzzy azimuth estimator. A trajectory controller for an omni-directional mobile robot, which each motor is controlled by an individual PID law to follow the speed command from inverse kinematics, needs a precise sensing data of its azimuth and exact estimation of reference azimuth value. It has imprecision and uncertainty inherent to perception sensors for azimuth. In this paper, they are solved by using fuzzy logic inference which can be used straightforward to perform the control of the mobile robot by means of the fuzzy behavior-based scheme already existent in literature. Finally, the good performance of the developed mobile robot is confirmed through live tests of path control task.

• Journal of KIISE:Databases
• /
• v.33 no.1
• /
• pp.138-149
• /
• 2006

As the environment with sensor network and ubiquitous computing is emerged, there are many demands of handling continuous, fast data such as streaming data. As work about streaming data has begun, work about management of streaming data in Publish-Subscribe system is started. The recent emergence of XML as a standard for information exchange on Internet has led to more interest in Publish - Subscribe system. A filtering technique of streaming XML data in the existing Publish- Subscribe system is using some schemes based on automata and YFilter, which is one of filtering techniques, is very popular. YFilter exploits commonality among path queries by sharing the common prefixes of the paths so that they are processed at most one and that is using the top-down approach. However, because partial matching path queries interrupt the common prefix sharing and don't calculate from root, throughput of YFilter decreases. So we use sharing of commonality among path queries with the common postfixes of the paths and use the bottom-up approach instead of the top-down approach. This filtering technique is called as PoSFilter. And we verify this technique through comparing with YFilter about throughput.

• The Transactions of the Korea Information Processing Society
• /
• v.7 no.11S
• /
• pp.3702-3713
• /
• 2000

Telecommunications standards have two different types of regulations: one is a law. enacted by government which all telecommunications related industries must observe. The other is a recommendatory standards. formulated by either government agency or some standardization organizations. Observation of these standards is not obligatory. However, technical standards are strict laws and ordinances based on common judgement and various conditions for evaluation of levels and limits. This paper deals with enhancing productivity of enactment and revision of technical standards. Through database of above related information we secure information continuity and public property of cyber space for the public. In this paper. we also classify recent data within the website in and out of the country offering four different methods of information retrieval and management system. The four retrieval methods suggested in this paper are itemized keyword retrieval. hierarchical retrieval, regulatory keyword retrieval and chronological keyword retrieval. These various retrieval methods provide the public with information of enactment and amendment of laws and regulations in the cyber space. thereby guarantees the sharing of information. Finally the important feature of the information retrieval system implemented in this paper is the online updating capability of law and regulations through the internet.

• Journal of Korea Spatial Information System Society
• /
• v.8 no.2 s.17
• /
• pp.91-107
• /
• 2006

To handling the extreme situation that must manage positional information of a very large volume, at least millions of moving objects. A cluster-based sealable distributed computing system architecture, called the GALIS which consists of multiple data processors, each dedicated to keeping records relevant to a different geographical zone and a different time zone, was proposed. In this paper, we proposed a valid time management and time-zone shifting scheme, which are essential in realizing the long-term location data subsystem of GALIS, but missed in our previous prototype development. We explain how to manage valid time of moving objects to avoid ambiguity of location information. We also describe time-zone shifting algorithm with three variations, such as Real Time-Time Zone Shifting, Batch-Time Zone Shifting, Table Partitioned Batch-Time Zone Shifting, Through experiments related with query processing time and CPU utilization, we show the efficiency of the proposed time-zone shifting schemes.

• Journal of Korea Spatial Information System Society
• /
• v.8 no.2 s.17
• /
• pp.75-89
• /
• 2006

LBS (Location-Based Service) systems have become a serious subject for research and development since recent rapid advances in wireless communication technologies and position measurement technologies such as global positioning system. The architecture named the GALIS (Gracefully Aging Location Information System) has been suggested which is a cluster-based distributed computing system architecture to overcome performance losses and to efficiently handle a large volume of data, at least millions. The GALIS consists of SLDS and LLDS. The SLDS manages current location information of moving objects and the LLDS manages past location information of moving objects. In this thesis, we implement a monitoring technique for the GALIS prototype, to allow dynamic load balancing among multiple computing nodes by keeping track of the load of each node in real-time during the location data management and spatio-temporal query processing. We also propose a buffering technique which efficiently manages the query results having overlapped query regions to improve query processing performance of the GALIS. The proposed scheme reduces query processing time by eliminating unnecessary query execution on the overlapped regions with the previous queries.

• Proceedings of the Korean Association of Geographic Inforamtion Studies Conference
• /
• 2010.06a
• /
• pp.185-185
• /
• 2010

문명 발전에 의해 생활공간이 더욱더 복잡해짐에 따라 현실 세계의 3차원 공간으로서의 이해가 필요해졌고, 3차원 공간을 효율적으로 관리 및 분석하기 위하여 3차원 GIS에 대한 중요성이 대두되기 시작하였다. 현재까지의 3차원 GIS는 1990년대에 들어서 3차원 가상도시의 등장으로 도시 모델의 3차원 가시화에 대해 많은 연구가 진행되어 왔다. 이러한 3차원 가상도시는 도시 관리 차원에서 도시공간의 표현, 분석, 모델링, 시뮬레이션 등을 통하여 도시정보의 체계적이고 효율적인 관리를 가능하게 한다. 3D GIS 기술의 활용 사례로는 3차원 도시 경관 분석 및 도시 계획, 3차원 시뮬레이션, 3차원 위치기반 서비스, 도시 행정 관리 등이 있다. 3차원 도시 경관 분석 및 도시 계획 활용 분야에서는 3차원 GIS를 활용하여 새롭게 건설되는 건물들이나 토지 이용 변경에 따른 경관 분석을 가상공간에서 시뮬레이션 분석함으로써 경관 변화 예측 및 적합한 경관 설계에 대한 분석을 한다. 3차원 시뮬레이션의 경우 환경오염도 분석, 최적의 입지선정, 개발 전후의 경관 차이, 각종 재난에 따른 위험도 분석, 경관과 일조량 분석 등을 미리 가상도시에 시뮬레이션 함으로서 불필요한 업무를 줄일 수 있고 발생할 수 있는 결과 값을 예상할 수 있다. 3차원 위치기반 서비스의 경우 자신의 위치를 비교적 정확히 측정함으로서 생소한 도시를 방문하거나 위치를 찾아갈 경우 최단 경로 및 최적 경로를 탐색할 수 있다. 앞으로의 3차원 GIS 기술 동향으로 유비쿼터스 환경에서의 시공간분석(Temporal GIS)과 3차원 실내공간인지 기술 개발 등으로 볼수있다. 시공간분석은 시공간 질의를 통한 공간정보의 시간에 다른 변화 계산 또는 예측된 결과를 분석할 수 있고, 시공간 분석에 사용되는 공간적, 시간적 정보의 조합은 적어도 3차원 이상의 정보를 나타내기 때문에 필요에 따라 다양한 결과를 산출 할 수 있다. 종합적으로 시공간분석은 GIS의 분석 범위를 넓혀 주며 실시간 GIS 응용시스템 개발을 위한 주요 과업이다. 3차원 실내공간정보 표현기법과 공간분석 및 공간인지 기법을 이용한 통한 유비쿼터스 환경에서의 3차원 실내공간인지 기술 개발을 통해 최근 공간정보를 기반으로 한 위치기반서비스내 u-서비스 즉, 인간 행태의 분석, 관리 및 응용을 지원하도록 한다 이는 위치기반서비스 사용자의 주변상황에 대한 인지능력을 향상시켜주고 최근 사회적 이슈가 되고 있는 U-City의 성공여부에 직접적인 영향을 미칠것이다. 이를 위해 다음과 같은 기술개발이 수행되어야 할것이다. 첫째로, 공간인지 기술을 위한 3차원 실내공간표현 방법 개발, 둘째로, 3차원 실내공간정보를 이용한 공간분석 및 공간인지 기법 개발, 마지막으로 유비쿼터스 환경에서 3차원 실내공간인지를 위한 미들웨어 개발을 통해서 공간정보의 활용을 macro-scale 공간에서 micro-scale로 확산될 것이다. 그러므로, 본 세미나를 통해서, 3차원 GIS의 활용의 보편화와 대중화를 위해서는 3차원 데이터 획득 및 처리 방법, 3차원 공간데이타베이스, 공간 분석 및 가시화에 대한 기술적 이슈들을 논하고자 한다.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.453-453
• /
• 2021

도시지역 비점오염원관리, 물순환 회복, 침투 및 증발산량 증가, 열섬현상 저감을 위한 주요한 방안으로 저영향개발(low impact development, LID)과 그린인프라 기법의 적용되고 있다. LID 시설은 소규모 분산형 시설로써 넓은 지역에 많고 다양한 시설들이 적용되어 시설의 개수가 많으며, 수질 및 토양 내 기성제품에 대한 센서들의 가격은 고가로 형성되어 있어 기기의 경제성 및 유지관리 등 적용하는데 제한적이다. 따라서 과거 모니터링 자료를 기반으로 오염물질들과의 상관성 분석을 통하여 계측이 어려운 항목들을 계측가능한 항목들로부터 예측 가능하며, 선정된 항목들에 대한 비용효율적인 센서를 개발하여 실시간 LID 모니터링이 가능한 비용효율적 모니터링을 개발하였다. 공주대학교 천안캠퍼스의 LID 시설들은 2013년에 조성되어 현재까지 시설이 운영되고 있으며, 5년이상의 과거 강우시 모니터링 자료들을 이용하여 오염물질 상관성 분석을 수행가능 하기에 대상지로 선정하였다. 오염물질 상관성 분석은 2013년부터 2017년도에 침투도랑에서 수행된 강우시 모니터링 자료를 활용하여 각 오염물질들의 상관성을 분석을 수행하였다. 침투도랑 내 유입되는 평균 유입수는 TSS 286.1±318.3 mg/L, BOD 22.6±39.5 mg/L, TN 8.96±5.85 mg/L, TP 1.01±1.11 mg/L로 나타났다. 겨울철에 비해 여름철에서의 오염물질의 유입농도가 높은 것으로 분석되었다. 이는 여름철 고온건조로 인한 노면 내 차량의 주행으로 인한 중금속, 폐타이어 등과 장마철 강우 시 유출된 토사로 인하여 유입수의 농도가 높은 것으로 분석되었다. 오염물질 부하량은 TSS와 COD 0.66으로 유의성이 높은 것으로 나왔으며, COD와 TSS, TP, TN 등 유의성이 높은 것으로 분석되었다. Arduino와 Raspberry PI를 활용하여 저비용 센서와 LTE 모뎀통신과 데이터 베이스 연결하여 개발된 프로그램을 통해서 무선으로 LID 시설에 대한 모니터링을 침투화분2와 식생체류지에 조성하였다. 전력공급이 어려운 식생체류지의 경우 태양열(Solar system) 시스템과 보조 전력 배터리를 조성하여 장마철이나 장기적인 악천후로 인한 전력을 생산하지 못할 경우 보조전력배터리에서 전력을 제공하여 지속적인 모니터링이 이루어지도록 설계하였다. 토양함수량, 토양온도와 Conductivity 등 3종류의 센서를 적용하였으며, 프로그램은 현재 2단계를 통한 2차수정을 통하여 프로그램을 구축하였다. 오차, 오작동, 계측값에 대한 검·보정 작업이 필요하다. 또한 대기자료의 구축을 통해 보다 토양과 LID 시설에 대한 영향분석이 필요한 것으로 사료된다.

• Journal of KIISE:Computing Practices and Letters
• /
• v.9 no.3
• /
• pp.322-331
• /
• 2003

As the complexity of a 3D game is increased by various factors of the game scenario, it has a problem for controlling the interrelation of the game objects. Therefore, a game system has a necessity of the coordination of the responses of the game objects. Also, it is necessary to control the behaviors of animations of the game objects in terms of the game scenario. To produce realistic game simulations, a system has to include a structure for designing the interactions among the game objects. This paper presents a method that designs the dynamic control mechanism for the interaction of the game objects in the game scenario. For the method, we suggest a game agent system as a framework that is based on intelligent agents who can make decisions using specific rules. Game agent systems are used in order to manage environment data, to simulate the game objects, to control interactions among game objects, and to support visual authoring interface that ran define a various interrelations of the game objects. These techniques can process the autonomy level of the game objects and the associated collision avoidance method, etc. Also, it is possible to make the coherent decision-making ability of the game objects about a change of the scene. In this paper, the rule-based behavior control was designed to guide the simulation of the game objects. The rules are pre-defined by the user using visual interface for designing their interaction. The Agent State Decision Network, which is composed of the visual elements, is able to pass the information and infers the current state of the game objects. All of such methods can monitor and check a variation of motion state between game objects in real time. Finally, we present a validation of the control method together with a simple case-study example. In this paper, we design and implement the supervised classification systems for high resolution satellite images. The systems support various interfaces and statistical data of training samples so that we can select the most effective training data. In addition, the efficient extension of new classification algorithms and satellite image formats are applied easily through the modularized systems. The classifiers are considered the characteristics of spectral bands from the selected training data. They provide various supervised classification algorithms which include Parallelepiped, Minimum distance, Mahalanobis distance, Maximum likelihood and Fuzzy theory. We used IKONOS images for the input and verified the systems for the classification of high resolution satellite images.