• 대한의용생체공학회:의공학회지
• /
• 제21권2호
• /
• pp.137-144
• /
• 2000

Clustered microcalcifications in X-ray mammograms are an important sign for the diagnosis of breast cancer. A shape-based method, which is based on the morphological features of clustered microcalcifications, is proposed for classifying clustered microcalcifications into benign or malignant categories. To verify the effectiveness of the proposed shape features, clinical mammograms were used to compare the classification performance of the proposed shape features with those of conventional textural features, such as the spatial gray-leve dependence method and the wavelet-based method. Image features extracted from these methods were used as inputs to a three-layer backpropagation neural network classifier. The classification performance of features extracted by each method was studied by using receiver operating-characteristics analysis. The proposed shape features were shown to be superior to the conventional textural features with respect to classification accuracy.

• 한국컴퓨터정보학회논문지
• /
• 제21권9호
• /
• pp.11-18
• /
• 2016

In this paper, we propose a gas classification method using combined features for an electronic nose system that performs well even when some loss occurs in measuring data samples. We first divide the entire measurement for a data sample into three local sections, which are the stabilization, exposure, and purge; local features are then extracted from each section. Based on the discrimination analysis, measurements of the discriminative information amounts are taken. Subsequently, the local features that have a large amount of discriminative information are chosen to compose the combined features together with the global features that extracted from the entire measurement section of the data sample. The experimental results show that the combined features by the proposed method gives better classification performance for a variety of volatile organic compound data than the other feature types, especially when there is data loss.

• Journal of Information Processing Systems
• /
• 제14권6호
• /
• pp.1318-1330
• /
• 2018

Video captioning refers to the process of extracting features from a video and generating video captions using the extracted features. This paper introduces a deep neural network model and its learning method for effective video captioning. In this study, visual features as well as semantic features, which effectively express the video, are also used. The visual features of the video are extracted using convolutional neural networks, such as C3D and ResNet, while the semantic features are extracted using a semantic feature extraction network proposed in this paper. Further, an attention-based caption generation network is proposed for effective generation of video captions using the extracted features. The performance and effectiveness of the proposed model is verified through various experiments using two large-scale video benchmarks such as the Microsoft Video Description (MSVD) and the Microsoft Research Video-To-Text (MSR-VTT).

• 마케팅과학연구
• /
• 제20권1호
• /
• pp.89-97
• /
• 2010

本文研究共同特性对于无选择权的影响, 并涉及到了调节焦点理论. 本文主要着眼于这三个因子以及他们之间的关系. 之前的研究已经广泛涉及到这三个方面. 第一, 共同特性影响已经被广为研究. Tversky (1972) 开创了这个理论, EBA 模型: 通过消除方面. 根据这个理论, 消费者在比较的过程中更易于注意特殊的特性, 而忽略共同特性. 最近, 更多的研究开始针对于此模型对于消费者行为的影响. Chernev (1997) 认为增加共同特性可以减少选择距离. 但是, 随后Chernev (2001) 的研究指出共同特性可能是消费者认知上的负担, 所以他们更喜欢启发式的过程而不是系统式的过程. 这些研究提出了一系列问题: 共同特性是否影响顾客选择? 如果是的话, 这些影响是什么样子的? 第二, 一些研究指出没有选择的状况是消费者最好的选择, 他们在犹豫不决时用这种方法回避选择. 其他关于这一理论的研究是时间的压力, 消费者自信, 以及可供选择的数量. 第三, 调节聚焦理论在目前非常流行. 消费者有两个焦点目标: 促进和制止. 促进聚焦主要和希望, 野心, 成功, 获得等有关; 而制止聚焦和责任, 职责, 安全, 规避等有关. 调节聚焦理论预测了顾客的感情, 创造, 态度, 记忆, 表现, 和判断. 而这些都是市场营销研究的领域这些文献为本文的研究提供了一些理论支持. 特别是增加共同特征而不是忽略他们可以增加选择过程中克制消费者的没有选择状况的比重, 其对于促进消费者的作用确实相反的. 本文通过两个试验进行验证. 第一个是2 X 2 组间的设计(共同特性X调节聚焦), 数码相机作为相关的客体. 特别的是, 调节聚焦变量是从11个问题中取得的. 共同特性包括焦距, 重量, 记忆卡, 电池, 而像素和价格作为独特特性. 结果证明了我们的假设, 那就是增加共同特性增加了克制消费者的无选择比重, 而对促进消费者没有作用. 第二个试验被用来复制第一个实验的结果. 这个实验和之前的基本相同, 只有两个方面不同—主要控制和研究客体. 在促进的前提下, 研究对象必须一些词例如: 利润, 野心, 高兴, 成功, 发展等. 在克制的前提下, 他们必学写下坚持, 安全, 保护, 规避, 损失, 责任等词. 实验证明我们假设是成立的. 本研究说明了共同特性对于顾客选择的二重效果. 增加共同特性可以提高或者降低无选择状况. 本文对于理论研究和实践上都有着贡献. 对于市场营销人员来说, 他们可能需要根据顾客的划分来考虑产品的共同特性. 理论上, 研究结果支持共同特性和无选择状况的调节变量. 最后, 本文也有一些不足, 例如过于强调态度的重要性等. 我们希望本文能够为未来的研究做出抛砖引玉的作用.

• 한국시뮬레이션학회논문지
• /
• 제20권4호
• /
• pp.13-20
• /
• 2011

현재 일상생활에서 얼굴 인식은 신원확인, 보안 등의 목적으로 사용되고 있다. 얼굴인식의 과정은 첫 번째로 얼굴이미지의 특징을 추출해야 한다. 다음으로 추출된 특징을 학습하고 그 중 학습이 잘된 식별력 있는 특징을 선택하게 된다. 그 이후 식별력 있는 특징을 이용하여 얼굴이미지를 인식하게 된다. 얼굴인식을 위해 사용하는 얼굴이미지의 특징의 수는 매우 많다. 이 많은 특징을 학습 및 인식에 다 사용할 경우 학습 시간과 컴퓨팅 자원의 효율성이 떨어지는 문제점을 가지고 있다. 이러한 문제를 해결하기 위해서 최근 여러가지의 Boosting 기법이 소개되어왔다. Boosting 기법은 특징을 효율적으로 선택하여 학습 알고리즘의 성능을 좋게해주는 기법이다. 그 중 MutualBoost라는 기법이 있는데 이 기법은 특징간의 상호정보를 이용하여 특징을 효율적으로 선택하게 하는 기법이다. 본 논문에서는 MutualBoost의 효과를 더 증대시키기 위해서 개별적인 특징학습이 아니라 특징들을 Group화하여 특징학습을 하는 GroupMutual-Boost기법을 제안한다. 특징들을 Group화 함으로써 특징의 학습 및 선택 시간이 줄어들게 되고 컴퓨팅 자원을 보다 효율적으로 사용할 수 있다.

• 패션비즈니스
• /
• 제8권4호
• /
• pp.33-44
• /
• 2004

In this study, based on the postmodern feminism were investigated in the non-fashion areas of painting, advertising and fashion areas. And, fashion illustrations were analyzed visually in the aspect of the essentialism and deconstructivism that constitute the postmodern feminism. In addition, it was examined how women images were expressed in fashion illustrations that reflected the postmodern feministic ideas. The research results first, in fashion illustrations were the essen tialistic women images were grouped in the opening of women bodies and actively emphasize. the opening of women bodies were expressed bodies the fetishistic, ecstatic images, actively emphasize features were expressed the sexual, and realistic images. Secondly, and fashion illustrations the deconstructivistic women images were grouped androgynous features, genderous features, the distorted feminine gender features, complex features. androgynous features were expressed the powerful, grotesque, humorous, androgynous image, that genderous features of immature, boyish image, that the distorted feminine gender features of simple, ethnic, techno-cyber image, that complex features of complex images.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2007년도 Proceedings of ISRS 2007
• /
• pp.14-17
• /
• 2007

To make use of surveying data obtained from different sensors and different techniques in a common reference frame, it is a pre-requite step to register them in a common coordinate system. For this purpose, we have developed a methodology to register IKONOS-2 Satellite Imagery using ALS data. To achieve this, conjugate features from these data should be extracted in advance. In the study, linear features are chosen as conjugate features because they can be accurately extracted from man-made structures in urban area, and more easily than point features from ALS data. Then, observation equations are established from similarity measurements of the extracted features. During the process, considering the characteristics of systematic errors in IKONOS-2 satellite imagery, the transformation function were selected and used. In addition, we also analyzed how the number of linear features and their spatial distribution used as control features affect the accuracy of registration. Finally, the results were evaluated statistically and the results clearly demonstrated that the proposed algorithms are appropriate to register these data.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제12권9호
• /
• pp.4412-4428
• /
• 2018

The purpose of this work is to solve the problem of representing an entire video using Convolutional Neural Network (CNN) features for human action recognition. Recently, due to insufficient GPU memory, it has been difficult to take the whole video as the input of the CNN for end-to-end learning. A typical method is to use sampled video frames as inputs and corresponding labels as supervision. One major issue of this popular approach is that the local samples may not contain the information indicated by the global labels and sufficient motion information. To address this issue, we propose a binary hashing method to enhance the local feature extractors. First, we extract the local features and aggregate them into global features using maximum/minimum pooling. Second, we use the binary hashing method to capture the motion features. Finally, we concatenate the hashing features with global features using different normalization methods to train the classifier. Experimental results on the JHMDB and MPII-Cooking datasets show that, for these new local features, binary hashing mapping on the sparsely sampled features led to significant performance improvements.

• 자동차안전학회지
• /
• 제8권2호
• /
• pp.6-10
• /
• 2016

The objective of this paper is to understand value of active safety features on the customer perspective. In this study, 30 participants who don't have experience with active safety features were recruited and asked for preference, usefulness and consideration of each active safety feature after driving evaluation by them. Through this research, the preference of active safety features were analyzed and which of active safety features were the most useful and the most considered by customer when they purchase new vehicle. As a result, adaptive cruise control and side blind zone alert were the most strongly preferred and considered features by respondents and it means that respondents wanted comfort environment while driving and seemed to value features that compensated for limited visibility. On the other hand, active safety features that warned driver without control of the vehicle was deemed generally less desirable such as lane departure warning and forward collision alert. But autonomous emergency braking was higher than the other active safety features with only warning even if they did not have experience for it while this test. They thought it will be helpful in case of front-end collision situation even they just listened description before the test.

• 수완나부미
• /
• 제2권2호
• /
• pp.53-68
• /
• 2010

In this paper, we examine the traditional culture features of Vietnam, Cambodia, and Thailand and above this we try to find out the evidences that traditional culture features are shown up through the Constitutions of three countries. First, for Vietnam, unlike the other two countries, we could find out the traditional culture feature of strong resistance and struggle against the foreign powers like China throughout the history. Also, the loyalty for the community and country could be added as the traditional culture features as well. In contrast, for Cambodia and Thailand, we could see other aspects of traditional culture features, such as the promotion and protection of Buddhism, historically sustained Kingship and respect for the King, the opportunistic foreign policy for the foreign powers to national survive. Furthermore, we could find out the evidences that these traditional culture features are shown implicitly or explicitly in the Constitutions of the three countries in all. Although three countries have different historical backgrounds, experiences, cultures, and value systems, we could see that the Constitution of the three countries show that the traditional culture features are vigorously alive in it. Thus, we can say that this proves that the flow of traditional culture features are not easily changed or stopped by an era or ideology and still alive in the heart of the people.