• Journal of Electrical Engineering and Technology
• /
• v.13 no.6
• /
• pp.2468-2478
• /
• 2018

We propose the Sparse Feature Convolutional Neural Network (SFCNN) to reduce the volume of convolutional neural networks (CNNs). Despite the superior classification performance of CNNs, their enormous network volume requires high computational cost and long processing time, making real-time applications such as online-training difficult. We propose an advanced network that reduces the volume of conventional CNNs by producing a region-based sparse feature map. To produce the sparse feature map, two complementary region-based value extraction methods, cluster max extraction and local value extraction, are proposed. Cluster max is selected as the main function based on experimental results. To evaluate SFCNN, we conduct an experiment with two conventional CNNs. The network trains 59 times faster and tests 81 times faster than the VGG network, with a 1.2% loss of accuracy in multi-class classification using the Caltech101 dataset. In vehicle classification using the GTI Vehicle Image Database, the network trains 88 times faster and tests 94 times faster than the conventional CNNs, with a 0.1% loss of accuracy.

• Journal of KIISE:Computer Systems and Theory
• /
• v.33 no.10
• /
• pp.768-776
• /
• 2006

Given an unstructured point set, we use an MLS (melting least-squares) approximation to estimate the local curvatures and their derivatives at a point by means of an approximation surface Then, we compute neighbor information using a Delaunay tessellation. feature points can then be detected as zero-crossings, and connected using curvature directions. Also this approach has a fast computation time than previous methods, which based on triangle meshes. We demonstrate our method on several large point-sampled models, rendered by point-splatting, on which the feature lines are rendered with line width determined from curvatures.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.9 no.7
• /
• pp.1513-1517
• /
• 2005

In this paper, I make use of a Multi-Channel skin color model with Hue, Cb, Cg using Red, Blue, Green channel altogether which remove bight component as being consider the characteristics of skin color to do modeling more effective to a facial skin color for extracting a facial area. 1 used efficient HOLA(Higher order local autocorrelation function) using 26 feature vectors to obtain both feature vectors of a facial area and the edge image extraction using Harr wavelet in image which split a facial area. Calculated feature vectors are used of date for the facial recognition through learning of neural network It demonstrate improvement in both the recognition rate and speed by proposed algorithm through simulation.

• ETRI Journal
• /
• v.38 no.1
• /
• pp.174-184
• /
• 2016

Information identification with image data by means of low-level visual features has evolved as a challenging research domain. Conventional text-based mapping of image data has been gradually replaced by content-based techniques of image identification. Feature extraction from image content plays a crucial role in facilitating content-based detection processes. In this paper, the authors have proposed four different techniques for multiview feature extraction from images. The efficiency of extracted feature vectors for content-based image classification and retrieval is evaluated by means of fusion-based and data standardization-based techniques. It is observed that the latter surpasses the former. The proposed methods outclass state-of-the-art techniques for content-based image identification and show an average increase in precision of 17.71% and 22.78% for classification and retrieval, respectively. Three public datasets - Wang; Oliva and Torralba (OT-Scene); and Corel - are used for verification purposes. The research findings are statistically validated by conducting a paired t-test.

• Journal of Communications and Networks
• /
• v.14 no.1
• /
• pp.21-26
• /
• 2012

As electrocardiogram (ECG) signals are generally sampled with a frequency of over 200 Hz, a method to compress diagnostic information without losing data is required to store and transmit them efficiently on a wireless personal area network (WPAN). In this paper, an ECG signal compression method for communications onWPAN, which uses feature points based on curvature, is proposed. The feature points of P, Q, R, S, and T waves, which are critical components of the ECG signal, have large curvature values compared to other vertexes. Thus, these vertexes were extracted with the proposed method, which uses local extrema of curvatures. Furthermore, in order to minimize reconstruction errors of the ECG signal, extra vertexes were added according to the iterative vertex selectionmethod. Through the experimental results on the ECG signals from Massachusetts Institute of Technology-Beth Israel hospital arrhythmia database, it was concluded that the vertexes selected by the proposed method preserved all feature points of the ECG signals. In addition, it was more efficient than the amplitude zone time epoch coding method.

• Journal of Korea Multimedia Society
• /
• v.20 no.5
• /
• pp.748-757
• /
• 2017

In order to recognize the facial expressions, good features that can express the facial expressions are essential. It is also essential to find the characteristic areas where facial expressions appear discriminatively. In this study, we propose a directional LBP feature for facial expression recognition and a method of finding directional LBP operation and feature region for facial expression classification. The proposed directional LBP features to characterize facial fine micro-patterns are defined by LBP operation factors (direction and size of operation mask) and feature regions through AdaBoost learning. The facial expression classifier is implemented as a SVM classifier based on learned discriminant region and directional LBP operation factors. In order to verify the validity of the proposed method, facial expression recognition performance was measured in terms of accuracy, sensitivity, and specificity. Experimental results show that the proposed directional LBP and its learning method are useful for facial expression recognition.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.48 no.4
• /
• pp.21-26
• /
• 2011

Feature ranking is useful to gain knowledge of data and identify relevant features. In this study, we proposed a use of feature ranking for classification of neuro-degeneration and vascular dementia in micro-Raman spectra of platelet. The entire region of the spectrum is divided into local region including several peaks, followed by Gaussian curve fitting method in the region to be modeled. Local minima select from the subregion and then remove the background based on the position by using interpolation method. After preprocessing steps, significant features were selected by feature ranking method to improve the classification accuracy and the computational complexity of classification system. PCA (principal component analysis) transform the selected features and the overall features that is used classification with the number of principal components. These were classified as MAP (maximum a posteriori) and it compared with classification result using overall features. In all experiments, the computational complexity of the classification system was remarkably reduced and the classification accuracy was partially increased. Particularly, the proposed method increased the classification accuracy in the experiment classifying the Parkinson's disease and normal with the average 1.7 %. From the result, it confirmed that proposed method could be efficiently used in the classification system of the neuro-degenerative disease and vascular dementia of platelet.

• The Journal of the Acoustical Society of Korea
• /
• v.14 no.2
• /
• pp.101-112
• /
• 1995

In order to construct a feature map-based phoneme classification system for speech recognition, two procedures are usually required. One is clustering and the other is labeling. In this paper, we present a phoneme classification system based on the Kohonen's Self-Organizing Feature Map (SOFM) for clusterer and labeler. It is known that the SOFM performs self-organizing process by which optimal local topographical mapping of the signal space and yields a reasonably high accuracy in recognition tasks. Consequently, SOFM can effectively be applied to the recognition of phonemes. Besides to improve the performance of the phoneme classification system, we propose the learning algorithm combined with the classical K-mans clustering algorithm in fine-tuning stage. In order to evaluate the performance of the proposed phoneme classification algorithm, we first use totaly 43 phonemes which construct six intra-class feature maps for six different phoneme classes. From the speaker-dependent phoneme classification tests using these six feature maps, we obtain recognition rate of $87.2\%$ and confirm that the proposed algorithm is an efficient method for improvement of recognition performance and convergence speed.

• Journal of Korea Entertainment Industry Association
• /
• v.13 no.6
• /
• pp.95-102
• /
• 2019

This paper mainly studies two films created by Huo Jianqi, Nuan and Postman in the Mountains, and analyzes the aesthetic style of Huo Jianqi's local movies in three parts. The first part deals with the characteristics of Huo Jianqi's local movies and discusses the theme selection, theme performance, character creation and emotional expression of the film. The second part elaborates Huo Jianqi's film and television works from three aspects: picture language, voice language and color language, which further reflects Huo Jianqi's aesthetic style. The third part analyses the puzzlement in the development of local films and how to develop them. With its unique oriental aesthetics, unique perspectives, unique rural scenery and characteristics, Chinese local movies have become the most representative of the overall level and style of Chinese movies. Chinese local film is the reflection and care of Chinese local culture, showing strong regional local culture, regionality is its distinct feature. Since the reform and opening up, with the changes of the times and the renewal of film directors, local films have shown distinct characteristics of the times. Through the analysis of Huo Jianqi's aesthetic style of local movies, this paper hopes to provide some valuable reference and inspiration for the development of Chinese local movies.

• Journal of the Korean Society for Precision Engineering
• /
• v.33 no.2
• /
• pp.129-138
• /
• 2016

This paper presents a natural corners-based SLAM (Simultaneous Localization and Mapping) with a robust data association algorithm in a real unknown environment. Corners are extracted from raw laser sensor data, which are chosen as landmarks for correcting the pose of mobile robot and building the map. In the proposed data association method, the extracted corners in every step are separated into several groups with small numbers of corners. In each group, local best matching vector between new corners and stored ones is found by joint compatibility, while nearest feature for every new corner is checked by individual compatibility. All these groups with local best matching vector and nearest feature candidate of each new corner are combined by partial compatibility with linear matching time. Finally, SLAM experiment results in an indoor environment based on the extracted corners show good robustness and low computation complexity of the proposed algorithms in comparison with existing methods.