• Title/Summary/Keyword: Performance Information Use

Search Result 5,643, Processing Time 0.045 seconds

Transfer Learning using Multiple ConvNet Layers Activation Features with Principal Component Analysis for Image Classification (전이학습 기반 다중 컨볼류션 신경망 레이어의 활성화 특징과 주성분 분석을 이용한 이미지 분류 방법)

  • Byambajav, Batkhuu;Alikhanov, Jumabek;Fang, Yang;Ko, Seunghyun;Jo, Geun Sik
    • Journal of Intelligence and Information Systems
    • /
    • v.24 no.1
    • /
    • pp.205-225
    • /
    • 2018
  • Convolutional Neural Network (ConvNet) is one class of the powerful Deep Neural Network that can analyze and learn hierarchies of visual features. Originally, first neural network (Neocognitron) was introduced in the 80s. At that time, the neural network was not broadly used in both industry and academic field by cause of large-scale dataset shortage and low computational power. However, after a few decades later in 2012, Krizhevsky made a breakthrough on ILSVRC-12 visual recognition competition using Convolutional Neural Network. That breakthrough revived people interest in the neural network. The success of Convolutional Neural Network is achieved with two main factors. First of them is the emergence of advanced hardware (GPUs) for sufficient parallel computation. Second is the availability of large-scale datasets such as ImageNet (ILSVRC) dataset for training. Unfortunately, many new domains are bottlenecked by these factors. For most domains, it is difficult and requires lots of effort to gather large-scale dataset to train a ConvNet. Moreover, even if we have a large-scale dataset, training ConvNet from scratch is required expensive resource and time-consuming. These two obstacles can be solved by using transfer learning. Transfer learning is a method for transferring the knowledge from a source domain to new domain. There are two major Transfer learning cases. First one is ConvNet as fixed feature extractor, and the second one is Fine-tune the ConvNet on a new dataset. In the first case, using pre-trained ConvNet (such as on ImageNet) to compute feed-forward activations of the image into the ConvNet and extract activation features from specific layers. In the second case, replacing and retraining the ConvNet classifier on the new dataset, then fine-tune the weights of the pre-trained network with the backpropagation. In this paper, we focus on using multiple ConvNet layers as a fixed feature extractor only. However, applying features with high dimensional complexity that is directly extracted from multiple ConvNet layers is still a challenging problem. We observe that features extracted from multiple ConvNet layers address the different characteristics of the image which means better representation could be obtained by finding the optimal combination of multiple ConvNet layers. Based on that observation, we propose to employ multiple ConvNet layer representations for transfer learning instead of a single ConvNet layer representation. Overall, our primary pipeline has three steps. Firstly, images from target task are given as input to ConvNet, then that image will be feed-forwarded into pre-trained AlexNet, and the activation features from three fully connected convolutional layers are extracted. Secondly, activation features of three ConvNet layers are concatenated to obtain multiple ConvNet layers representation because it will gain more information about an image. When three fully connected layer features concatenated, the occurring image representation would have 9192 (4096+4096+1000) dimension features. However, features extracted from multiple ConvNet layers are redundant and noisy since they are extracted from the same ConvNet. Thus, a third step, we will use Principal Component Analysis (PCA) to select salient features before the training phase. When salient features are obtained, the classifier can classify image more accurately, and the performance of transfer learning can be improved. To evaluate proposed method, experiments are conducted in three standard datasets (Caltech-256, VOC07, and SUN397) to compare multiple ConvNet layer representations against single ConvNet layer representation by using PCA for feature selection and dimension reduction. Our experiments demonstrated the importance of feature selection for multiple ConvNet layer representation. Moreover, our proposed approach achieved 75.6% accuracy compared to 73.9% accuracy achieved by FC7 layer on the Caltech-256 dataset, 73.1% accuracy compared to 69.2% accuracy achieved by FC8 layer on the VOC07 dataset, 52.2% accuracy compared to 48.7% accuracy achieved by FC7 layer on the SUN397 dataset. We also showed that our proposed approach achieved superior performance, 2.8%, 2.1% and 3.1% accuracy improvement on Caltech-256, VOC07, and SUN397 dataset respectively compare to existing work.

A 10b 50MS/s Low-Power Skinny-Type 0.13um CMOS ADC for CIS Applications (CIS 응용을 위해 제한된 폭을 가지는 10비트 50MS/s 저 전력 0.13um CMOS ADC)

  • Song, Jung-Eun;Hwang, Dong-Hyun;Hwang, Won-Seok;Kim, Kwang-Soo;Lee, Seung-Hoon
    • Journal of the Institute of Electronics Engineers of Korea SD
    • /
    • v.48 no.5
    • /
    • pp.25-33
    • /
    • 2011
  • This work proposes a skinny-type 10b 50MS/s 0.13um CMOS three-step pipeline ADC for CIS applications. Analog circuits for CIS applications commonly employ a high supply voltage to acquire a sufficiently acceptable dynamic range, while digital circuits use a low supply voltage to minimize power consumption. The proposed ADC converts analog signals in a wide-swing range to low voltage-based digital data using both of the two supply voltages. An op-amp sharing technique employed in residue amplifiers properly controls currents depending on the amplification mode of each pipeline stage, optimizes the performance of op-amps, and improves the power efficiency. In three FLASH ADCs, the number of input stages are reduced in half by the interpolation technique while each comparator consists of only a latch with low kick-back noise based on pull-down switches to separate the input nodes and output nodes. Reference circuits achieve a required settling time only with on-chip low-power drivers and digital correction logic has two kinds of level shifter depending on signal-voltage levels to be processed. The prototype ADC in a 0.13um CMOS to support 0.35um thick-gate-oxide transistors demonstrates the measured DNL and INL within 0.42LSB and 1.19LSB, respectively. The ADC shows a maximum SNDR of 55.4dB and a maximum SFDR of 68.7dB at 50MS/s, respectively. The ADC with an active die area of 0.53$mm^2$ consumes 15.6mW at 50MS/s with an analog voltage of 2.0V and two digital voltages of 2.8V ($=D_H$) and 1.2V ($=D_L$).

A Study on the Method of Producing the 1 km Resolution Seasonal Prediction of Temperature Over South Korea for Boreal Winter Using Genetic Algorithm and Global Elevation Data Based on Remote Sensing (위성고도자료와 유전자 알고리즘을 이용한 남한의 겨울철 기온의 1 km 격자형 계절예측자료 생산 기법 연구)

  • Lee, Joonlee;Ahn, Joong-Bae;Jung, Myung-Pyo;Shim, Kyo-Moon
    • Korean Journal of Remote Sensing
    • /
    • v.33 no.5_2
    • /
    • pp.661-676
    • /
    • 2017
  • This study suggests a new method not only to produce the 1 km-resolution seasonal prediction but also to improve the seasonal prediction skill of temperature over South Korea. This method consists of four stages of experiments. The first stage, EXP1, is a low-resolution seasonal prediction of temperature obtained from Pusan National University Coupled General Circulation Model, and EXP2 is to produce 1 km-resolution seasonal prediction of temperature over South Korea by applying statistical downscaling to the results of EXP1. EXP3 is a seasonal prediction which considers the effect of temperature changes according to the altitude on the result of EXP2. Here, we use altitude information from ASTER GDEM, satellite observation. EXP4 is a bias corrected seasonal prediction using genetic algorithm in EXP3. EXP1 and EXP2 show poorer prediction skill than other experiments because the topographical characteristic of South Korea is not considered at all. Especially, the prediction skills of two experiments are lower at the high altitude observation site. On the other hand, EXP3 and EXP4 applying the high resolution elevation data based on remote sensing have higher prediction skill than other experiments by effectively reflecting the topographical characteristics such as temperature decrease as altitude increases. In addition, EXP4 reduced the systematic bias of seasonal prediction using genetic algorithm shows the superior performance for temporal variability such as temporal correlation, normalized standard deviation, hit rate and false alarm rate. It means that the method proposed in this study can produces high-resolution and high-quality seasonal prediction effectively.

Wavelet Transform-based Face Detection for Real-time Applications (실시간 응용을 위한 웨이블릿 변환 기반의 얼굴 검출)

  • 송해진;고병철;변혜란
    • Journal of KIISE:Software and Applications
    • /
    • v.30 no.9
    • /
    • pp.829-842
    • /
    • 2003
  • In this Paper, we propose the new face detection and tracking method based on template matching for real-time applications such as, teleconference, telecommunication, front stage of surveillance system using face recognition, and video-phone applications. Since the main purpose of paper is to track a face regardless of various environments, we use template-based face tracking method. To generate robust face templates, we apply wavelet transform to the average face image and extract three types of wavelet template from transformed low-resolution average face. However template matching is generally sensitive to the change of illumination conditions, we apply Min-max normalization with histogram equalization according to the variation of intensity. Tracking method is also applied to reduce the computation time and predict precise face candidate region. Finally, facial components are also detected and from the relative distance of two eyes, we estimate the size of facial ellipse.

Business Strategies for Korean Private Security-Guard Companies Utilizing Resource-based Theory and AHP Method (자원기반 이론과 AHP 방법을 활용한 민간 경호경비 기업의 전략 연구)

  • Kim, Heung-Ki;Lee, Jong-Won
    • Korean Security Journal
    • /
    • no.36
    • /
    • pp.177-200
    • /
    • 2013
  • As we enter a high industrial society that widens the gap between the rich and poor, demand for the security services has grown explosively. With the growth in quantitative expansion of security services, people have also placed increased requirements on more sophisticated and diversified security services. Consequently, market outlook for private security services industry is positive. However, Korea's private security services companies are experiencing difficulties in finding a direction to capture this new market opportunity due to their small sizes and lack of management-strategic thinking skills. Therefore, we intend to offer a direction of development for our private security services industry using a management-strategy theory and the Analytic Hierarchy Process(AHP), a structured decision-making method. A resource-based theory is one of the important management strategy theories. It explains that a company's overall performance is primarily determined by its competitive resources. Using this theory, we could analyze a company's unique resources and core competencies and set a strategic direction for the company accordingly. The usefulness and validity of this theory has been demonstrated as it has often been subject to empirical verification since 1990s. Based on this theory, we outlined a set of basic procedures to establish a management strategy for the private security services companies. We also used the AHP method to identify competitive resources, core competencies, and strategies from private security services companies in contrast with public companies. The AHP method is a technique that can be used in the decision making process by quantifying experts' knowledge and unstructured problems. This is a verified method that has been used in the management decision making in the corporate environment as well as for the various academic studies. In order to perform this method, we gathered data from 11 experts from academic, industrial, and research sectors and drew distinctive resources, competencies, and strategic direction for private security services companies vis-a-vis public organizations. Through this process, we came to the conclusion that private security services companies generally have intangible resources as their distinctive resources compared with public organization. Among those intangible resources, relational resources, customer information, and technologies were analyzed as important. In contrast, tangible resources such as equipment, funds, distribution channels are found to be relatively scarce. We also found the competencies in sales and marketing and new product development as core competencies. We chose a concentration strategy focusing on a particular market segment as a strategic direction considering these resources and competencies of private security services companies. A concentration strategy is the right fit for smaller companies as a strategy to allow them to focus all of their efforts on target customers in a single segment. Thus, private security services companies would face the important tasks such as developing a new market and appropriate products for such market segment and continuing marketing activities to manage their customers. Additionally, continuous recruitment is required to facilitate the effective use of human resources in order to strengthen their marketing competency in a long term.

  • PDF

Acoustic images of the submarine fan system of the northern Kumano Basin obtained during the experimental dives of the Deep Sea AUV URASHIMA (심해 자율무인잠수정 우라시마의 잠항시험에서 취득된 북 구마노 분지 해저 선상지 시스템의 음향 영상)

  • Kasaya, Takafumi;Kanamatsu, Toshiya;Sawa, Takao;Kinosita, Masataka;Tukioka, Satoshi;Yamamoto, Fujio
    • Geophysics and Geophysical Exploration
    • /
    • v.14 no.1
    • /
    • pp.80-87
    • /
    • 2011
  • Autonomous underwater vehicles (AUVs) present the important advantage of being able to approach the seafloor more closely than surface vessel surveys can. To collect bathymetric data, bottom material information, and sub-surface images, multibeam echosounder, sidescan sonar (SSS) and subbottom profiler (SBP) equipment mounted on an AUV are powerful tools. The 3000m class AUV URASHIMA was developed by the Japan Agency for Marine-Earth Science and Technology (JAMSTEC). After finishing the engineering development and examination phase of a fuel-cell system used for the vehicle's power supply system, a renovated lithium-ion battery power system was installed in URASHIMA. The AUV was redeployed from its prior engineering tasks to scientific use. Various scientific instruments were loaded on the vehicle, and experimental dives for science-oriented missions conducted from 2006. During the experimental cruise of 2007, high-resolution acoustic images were obtained by SSS and SBP on the URASHIMA around the northern Kumano Basin off Japan's Kii Peninsula. The map of backscatter intensity data revealed many debris objects, and SBP images revealed the subsurface structure around the north-eastern end of our study area. These features suggest a structure related to the formation of the latest submarine fan. However, a strong reflection layer exists below ~20 ms below the seafloor in the south-western area, which we interpret as a denudation feature, now covered with younger surface sediments. We continue to improve the vehicle's performance, and expect that many fruitful results will be obtained using URASHIMA.

Criteria of Evaluating Clothing and Web Service on Internet Shopping Mall Related to Consumer Involvement (인터넷 쇼핑몰 이용자의 소비자 관여에 따른 의류제품 및 웹 서비스 평가기준에 관한 연구)

  • Lee, Kyung-Hoon;Park, Jae-Ok
    • Journal of the Korean Society of Clothing and Textiles
    • /
    • v.30 no.12 s.159
    • /
    • pp.1747-1758
    • /
    • 2006
  • Rapid development of the information technology has influenced on the changes in every sector of human environments. One prominent change in retail market is an increase of electronic stores, which has prompted practical and research interest in the product and store attributes that include consumer to purchase products from the electronic shopping. Therefore many marketers are paying much attention to the criteria of evaluating clothing and web service on internet shopping malls. The purpose of this study is to examine differences of clothing and web service criteria of consumer groups (High-Involvement & High-Ability, Low-Involvement & High-Ability, High-Involvement & Low-Ability, and Low-Involvement & Low-Ability) who are classified into consumer involvement and internet use ability. The subjects of this study were 305 people aged between 19 and 39s, living in Seoul and Gyeonggi-do area, and having experiences in buying products on the internet shopping. Statistical analyses used for this study were the frequency, percentage, factor analysis, ANOVA and Duncan test. The results of this study were as follows: Regarded on the criteria of evaluating clothing, the low different groups had significant differences in the esthetic, the quality performance and the extrinsic criterion. Both HIHA group and HILA group showed the similar results. They considered every criterion of evaluating clothing more important, compared with other groups. Regarded on the criteria of evaluating web service related to the low different groups, there were significant differences in the factors related to the shopping mall reliance, the product, the satisfaction after purchase, and the promotion and policy criterion. Both HIHA group and HILA group showed the similar results as well. They considered every criterion of evaluating web service more important, compared with other groups. In conclusion, HI groups perceive relatively more dangerous factors which can be occurred during internet shopping. Therefore, internet shopping malls need to provide clothing that can satisfy the HI groups as well as make efforts to remove the dangerous factors on the internet.

T-Cache: a Fast Cache Manager for Pipeline Time-Series Data (T-Cache: 시계열 배관 데이타를 위한 고성능 캐시 관리자)

  • Shin, Je-Yong;Lee, Jin-Soo;Kim, Won-Sik;Kim, Seon-Hyo;Yoon, Min-A;Han, Wook-Shin;Jung, Soon-Ki;Park, Se-Young
    • Journal of KIISE:Computing Practices and Letters
    • /
    • v.13 no.5
    • /
    • pp.293-299
    • /
    • 2007
  • Intelligent pipeline inspection gauges (PIGs) are inspection vehicles that move along within a (gas or oil) pipeline and acquire signals (also called sensor data) from their surrounding rings of sensors. By analyzing the signals captured in intelligent PIGs, we can detect pipeline defects, such as holes and curvatures and other potential causes of gas explosions. There are two major data access patterns apparent when an analyzer accesses the pipeline signal data. The first is a sequential pattern where an analyst reads the sensor data one time only in a sequential fashion. The second is the repetitive pattern where an analyzer repeatedly reads the signal data within a fixed range; this is the dominant pattern in analyzing the signal data. The existing PIG software reads signal data directly from the server at every user#s request, requiring network transfer and disk access cost. It works well only for the sequential pattern, but not for the more dominant repetitive pattern. This problem becomes very serious in a client/server environment where several analysts analyze the signal data concurrently. To tackle this problem, we devise a fast in-memory cache manager, called T-Cache, by considering pipeline sensor data as multiple time-series data and by efficiently caching the time-series data at T-Cache. To the best of the authors# knowledge, this is the first research on caching pipeline signals on the client-side. We propose a new concept of the signal cache line as a caching unit, which is a set of time-series signal data for a fixed distance. We also provide the various data structures including smart cursors and algorithms used in T-Cache. Experimental results show that T-Cache performs much better for the repetitive pattern in terms of disk I/Os and the elapsed time. Even with the sequential pattern, T-Cache shows almost the same performance as a system that does not use any caching, indicating the caching overhead in T-Cache is negligible.

A Phenomenological Study for Hospitalized Elderly무s Powerlessness (병원에 입원한 노인의 무력감 현상 연구)

  • 최영희;김경은
    • Journal of Korean Academy of Nursing
    • /
    • v.26 no.1
    • /
    • pp.223-247
    • /
    • 1996
  • This study was done to provide information which would lead to nursing care of the elderly being more holistically through an understanding of the phenomena of powerlessness based on the lived experience of powerlessness by the elderly, the meaning the elderly give to such phenomena, and what essence of powerlessness is. The methodology used in this study was Max Van Manen's phenomenological method based on the philosophy of Merleu-Ponty and a concerted approach was realized through the 11 steps suggested in the Van Manen's method. Data collection was done from March 2, 1995 to December 30, 1995. The subjects for this study were four elderly persons who lived with their families and who were over 60 years of age. Data were collected about the lived experience of the elderly, this researcher's experience of powerlessness, the linguistic meaning of powerlessness, idioms of the word or a feeling of powerlessness, and descriptions of powerlessness in the elderly as they appeared in the literature, are works, and phenomenological literature. All data were used to provide insights into the phenomena of powerlessness. Data about the experience of powerlessness by the elderly were collected through open interviews, participation, and observation. In the analysis of the theme of this study, the aspects of the theme, powerlessness in the elderly were clarified, thereby abstracting and finding meaningful statements by the elderly about their feeling of powerlessness, and then those significant statements were expressed as linguistic transformations. The summarized findings from the study are as follows : 1. Five meanings of powerlessness in the elderly were defined. 〈weakness〉, 〈dependence〉, 〈frustration〉, 〈worthlessness〉 and 〈giving up〉. 2. 〈Weakness〉 means that the elderly experience, not only their aging but also, their becoming weak and the loss of physical function frequently caused by diseases. 〈Dependence〉 means that the elderly experience dependence without any influence from the surroundings and that elderly patients who are hospitalized lose their autonomy, follow entirely their doctor's prescriptions, use aid equipment and directions, and depend only on those things. 〈Frustration〉 means that the elderly experience the loss of their roles from the past, there by feeling that there is no work for them to do anymore and therefore feel unable to do anything. 〈Worthlessness〉 means that the elderly experience the feeling of losing their social roles from the past, having no financial ability, thereby being a burden to their children or the people around them, and therefore regarding themselves useless. 〈Giving up〉 means that the elderly experience the feeling of closeness to death in the final stage of their lifetime, lose hope to be healed from their disease, and recognize the incontrollability of their own body. 3. From a general view of the meaning of the theme the powerlessness in the elderly-the most essential meaning of the theme is the 〈sense of loss〉. For the elderly are experiencing a sense of loss in the situation of being elderly and therefore being often hospitalized. Brief definitions of the five phenomena could be 〈weakness〉 meaning the loss of physical strength, 〈dependence〉 the loss of mentality caused by disease and hospitalization, 〈frustration〉 and 〈worthlessness〉 the loss of social performance caused by the loss of social functions from the past, and lastly 〈giving up〉 the loss of the controllability of such situations of aging and suffering disease. In light of the discussion above, it is understandable that the hospitalized elderly experience powerlessness not only as it related to their diseases but also to their normal aging, and this related to other characteristics of being elderly means that the 〈sense of loss〉 is the very essence of their powerlessness. 4. While most cases are of the normal elderly experiencing powerlessness in relation to their social network, cases of elderly who are hospitalized are of those experiencing powerlessness in relation to the loss of their physical desire. 5. The findings discussed above can serve as guidelines for nurses who take care of the ill elderly who are hospitalized and that can provide cues to appropriate nursing service, recognizing that the subjective experience of the objective age of the elderly is so important. Nurses can provide highly qualitative nursing service, based on their deep understanding of the suffering of the elderly due to feelings of powerlessness.

  • PDF

A Study on Establishment of Safety Training Center Based on Virtual Reality and Augmented Reality Technology for Military Safety and Suicide Accident Prevention (가상현실(VR/AR) 기술 기반으로 군 안전 및 자살사고 예방을 위한 안전체험훈련장 구축 방안에 관한 연구)

  • Choi, Sung-oh;Min, Yong-sik;Kim, Sung-Il;Ghoi, Jong-geun
    • Journal of Internet Computing and Services
    • /
    • v.21 no.2
    • /
    • pp.139-148
    • /
    • 2020
  • Due to change in circumstances in the 2000s such as severe birthrate decline and shortened military service period, the armed forces of the Republic of Korea is currently turning to technologies and equipments from manpower, developing it to become high-tech, high-speed, and complex, resulting in an environment in which a single mistake could cause a mass mortality crisis.It is also evident that, considering aspects such as safety training curriculums and achievements of advanced countries and private education, hands-on training is a must in preventing suicides and accidents in the military, and establishing safety training centers is crucial for systematic and effective hands-on training.Soldiers who are joining the army as of now have experienced the Internet ever since they were born and easily use both virtual and augmented reality, and the current level of science and technology has developed to the point where most of the public safety experience centers are able to be replaced by virtual and augmented reality (VR/AR). Therefore, considering the aspects such as installation space, construction costs, maintenance costs, user characteristics, and education effects, other than for those trainings where real models and objects are more effective such as first aid training, it is with a strong recommendation that establishing military safety training facilities with VR/AR (Virtual and augmented reality) is a must in the coming future. We have derived the need for hands-on training by considering the development of virtual and augmented reality (VR/AR), analysis of operation status of the public safety experience centers, characteristics of military units, installation and maintenance costs, and proposed plan to establish safety training centers where effective training performance can be achieved at a lower cost than the public safety experience center. In addition, we suggested the scale of the required safety training center and the composition of the experience rooms considering the number of trainees and the environment of each military units. Given this analysis it will contribute to the prevention of military safety and suicide by building a safety training center in the future.