• 한국멀티미디어학회논문지
• /
• 제11권6호
• /
• pp.807-815
• /
• 2008

A simple efficient method of spotting and recognizing hand gestures in video is presented using a network of hidden Markov models and dynamic programming search algorithm. The description starts from designing a set of isolated trajectory models which are stochastic and robust enough to characterize highly variable patterns like human motion, handwriting, and speech. Those models are interconnected to form a single big network termed a spotting network or a spotter that models a continuous stream of gestures and non-gestures as well. The inference over the model is based on dynamic programming. The proposed model is highly efficient and can readily be extended to a variety of recurrent pattern recognition tasks. The test result without any engineering has shown the potential for practical application. At the end of the paper we add some related experimental result that has been obtained using a different model - dynamic Bayesian network - which is also a type of stochastic model.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2005년도 학술발표회(1)
• /
• pp.16-16
• /
• 2005

• 방송공학회논문지
• /
• 제26권4호
• /
• pp.368-376
• /
• 2021

인공지능 클라우드는 학습된 모델 공유 및 실행 환경을 제공하여 인공지능 기술과 제어 기술을 융합하는 자율 사물 개발을 지원한다. 기존 자율 사물 개발 기술은 인공지능 모델의 정확도만을 고려하여 은닉 계층 수 및 커널 수 증가 등 모델의 복잡성을 증가시켜 결과적으로 많은 연산량을 요구하게 한다. 자원 제약적 컴퓨팅 환경은 해당 모델이 필요로 하는 충분한 자원을 제공할 수 없어 자율 사물의 실시간성 장애를 발생시킬 수 있다. 본 논문은 컴퓨팅 환경에 최적화된 인공지능 모델을 선택하는 디지털 트윈 소프트웨어 개발 프레임워크를 제안한다. 제안 프레임워크는 DNN 기반 부하 예측 모델을 활용하여 제어 소프트웨어를 개발한다. 부하 예측 모델은 디지털 트윈을 활용하여 인공지능 모델의 부하를 예측하여 특정 컴퓨팅 환경에 최적의 모델 선택을 지원한다. 대표적인 CNN 모델을 활용한 부하 예측 실험으로 제안 부하 예측 DNN 모델이 수식 기반 부하 예측 대비 최대 20%의 오류를 보임을 확인했다.

• 산업경영시스템학회지
• /
• 제46권1호
• /
• pp.55-67
• /
• 2023

Lightweight face recognition models, as one of the most popular and long-standing topics in the field of computer vision, has achieved vigorous development and has been widely used in many real-world applications due to fewer number of parameters, lower floating-point operations, and smaller model size. However, few surveys reviewed lightweight models and reimplemented these lightweight models by using the same calculating resource and training dataset. In this survey article, we present a comprehensive review about the recent research advances on the end-to-end efficient lightweight face recognition models and reimplement several of the most popular models. To start with, we introduce the overview of face recognition with lightweight models. Then, based on the construction of models, we categorize the lightweight models into: (1) artificially designing lightweight FR models, (2) pruned models to face recognition, (3) efficient automatic neural network architecture design based on neural architecture searching, (4) Knowledge distillation and (5) low-rank decomposition. As an example, we also introduce the SqueezeFaceNet and EfficientFaceNet by pruning SqueezeNet and EfficientNet. Additionally, we reimplement and present a detailed performance comparison of different lightweight models on the nine different test benchmarks. At last, the challenges and future works are provided. There are three main contributions in our survey: firstly, the categorized lightweight models can be conveniently identified so that we can explore new lightweight models for face recognition; secondly, the comprehensive performance comparisons are carried out so that ones can choose models when a state-of-the-art end-to-end face recognition system is deployed on mobile devices; thirdly, the challenges and future trends are stated to inspire our future works.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제10권6호
• /
• pp.2767-2780
• /
• 2016

Video-based human-activity recognition has become increasingly popular due to the prominent corresponding applications in a variety of fields such as computer vision, image processing, smart-home healthcare, and human-computer interactions. The essential goals of a video-based activity-recognition system include the provision of behavior-based information to enable functionality that proactively assists a person with his/her tasks. The target of this work is the development of a novel approach for human-activity recognition, whereby human-body-joint features that are extracted from depth videos are used. From silhouette images taken at every depth, the direction and magnitude features are first obtained from each connected body-joint pair so that they can be augmented later with motion direction, as well as with the magnitude features of each joint in the next frame. A generalized discriminant analysis (GDA) is applied to make the spatiotemporal features more robust, followed by the feeding of the time-sequence features into a Hidden Markov Model (HMM) for the training of each activity. Lastly, all of the trained-activity HMMs are used for depth-video activity recognition.

• 정보교육학회논문지
• /
• 제9권3호
• /
• pp.407-416
• /
• 2005

컴퓨터 교육을 위한 교육방법의 논의는 학습자가 컴퓨터에 대한 지식과 기능을 효과적으로 습득하기 위해 요구되는 사항이다. 본 논문에서는 다양한 교수법 중 교사와 학습자간의 부담감을 줄이기 위해 STAD협동학습을 적용해 보았다. 그 결과 STAD협동학습 이론에 기초한 컴퓨터 실기 수업이 학생들의 ICT 정보수집 및 정보가공 능력 향상에 효과적인임을 알 수 있었다. 연구결과를 통해 도출된 수업모형은 컴퓨터 과의 수업모형 중 하나의 예시가 될 것이며, 앞으로 컴퓨터 교과의 효과적인 수업을 위한 더 많은 연구사례가 있어야 한다고 생각한다.

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 2011년도 한국컴퓨터종합학술대회논문집 Vol.38 No.1(A)
• /
• pp.112-115
• /
• 2011

Healthcare is one of the biggest industries that consume IT Infrastructure as a necessity. A reliable reach to secure private information is the essential goal. Most of the healthcare organizations either have their own custom solutions developed or third part CRM software tailored to their needs. This culture satisfies the usability internal to the organization but lacks the ability to provide public access to the users or even to take a step ahead and collaborate with like minded institutes for research and potential growth. Emergence of cloud computing has provided us the opportunity to take a step ahead and build applications available to users on a secure platform, with ability to scale the resources depending on the requirements. The effective use of hybrid and community cloud can create rock solid foundations for healthcare IT Infrastructure. These cloud deployment models not only caters the present needs of healthcare IT Infrastructure but also opens the possibilities to build applications and provide secure and reliable access beyond the limits of organizational boundaries.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제10권3호
• /
• pp.1390-1403
• /
• 2016

Facial expression recognition (FER) plays a very significant role in computer vision, pattern recognition, and image processing applications such as human computer interaction as it provides sufficient information about emotions of people. For video-based facial expression recognition, depth cameras can be better candidates over RGB cameras as a person's face cannot be easily recognized from distance-based depth videos hence depth cameras also resolve some privacy issues that can arise using RGB faces. A good FER system is very much reliant on the extraction of robust features as well as recognition engine. In this work, an efficient novel approach is proposed to recognize some facial expressions from time-sequential depth videos. First of all, efficient Local Binary Pattern (LBP) features are obtained from the time-sequential depth faces that are further classified by Generalized Discriminant Analysis (GDA) to make the features more robust and finally, the LBP-GDA features are fed into Hidden Markov Models (HMMs) to train and recognize different facial expressions successfully. The depth information-based proposed facial expression recognition approach is compared to the conventional approaches such as Principal Component Analysis (PCA), Independent Component Analysis (ICA), and Linear Discriminant Analysis (LDA) where the proposed one outperforms others by obtaining better recognition rates.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제7권3호
• /
• pp.558-575
• /
• 2013

The purpose of a data leakage prevention system is to protect corporate information assets. The system monitors the packet exchanges between internal systems and the Internet, filters packets according to the data security policy defined by each company, or discretionarily deletes important data included in packets in order to prevent leakage of corporate information. However, the problem arises that the system may monitor employees' personal information, thus allowing their privacy to be violated. Therefore, it is necessary to find not only a solution for detecting leakage of significant information, but also a way to minimize the leakage of internal users' personal information. In this paper, we propose two models for representing the level of personal information disclosure during data leakage detection. One model measures only the disclosure frequencies of keywords that are defined as personal data. These frequencies are used to indicate the privacy violation level. The other model represents the context of privacy violation using a private data matrix. Each row of the matrix represents the disclosure counts for personal data keywords in a given time period, and each column represents the disclosure count of a certain keyword during the entire observation interval. Using the suggested matrix model, we can represent an abstracted context of the privacy violation situation. Experiments on the privacy violation situation to demonstrate the usability of the suggested models are also presented.

• International Journal of Control, Automation, and Systems
• /
• 제4권3호
• /
• pp.372-381
• /
• 2006

The purpose of this study was to develop a process management system to manage ingot fabrication and improve ingot quality. The ingot is the first manufactured material of wafers. Trace parameters were collected on-line but measurement parameters were measured by sampling inspection. The quality parameters were applied to evaluate the quality. Therefore, preprocessing was necessary to extract useful information from the quality data. First, statistical methods were used for data generation. Then, modeling was performed, using the generated data, to improve the performance of the models. The function of the models is to predict the quality corresponding to control parameters. Secondly, rule extraction was performed to find the relation between the production quality and control conditions. The extracted rules can give important information concerning how to handle the process correctly. The dynamic polynomial neural network (DPNN) and decision tree were applied for data modeling and rule extraction, respectively, from the ingot fabrication data.