• The Journal of the Acoustical Society of Korea
• /
• v.3 no.1
• /
• pp.25-35
• /
• 1984

This paper analysis the Dynamic Time Warping algorithms for time normalization of speech pattern and discusses the Dynamic Programming algorithm for spoken Korean isolated digits recognition. In the DP matching, feature vectors of the reference and test pattern are consisted of first three formant frequencies extracted by power spectrum density estimation algorithm of the ARMA model. The major differences in the various DTW algorithms include the global path constrains, the local continuity constraints on the path, and the distance weighting/normalization used to give the overall minimum distance. The performance criterias to evaluate these DP algorithms are memory requirement, speed of implementation, and recognition accuracy.

• Journal of Korea Multimedia Society
• /
• v.23 no.8
• /
• pp.1019-1029
• /
• 2020

In this paper, we propose a novel knowledge distillation algorithm to create an compressed deep ensemble network coupled with the combined use of local and global features of face images. In order to transfer the capability of high-level recognition performances of the ensemble deep networks to a single deep network, the probability for class prediction, which is the softmax output of the ensemble network, is used as soft target for training a single deep network. By applying the knowledge distillation algorithm, the local feature informations obtained by training the deep ensemble network using facial subregions of the face image as input are transmitted to a single deep network to create a so-called compressed ensemble DCNN. The experimental results demonstrate that our proposed compressed ensemble deep network can maintain the recognition performance of the complex ensemble deep networks and is superior to the recognition performance of a single deep network. In addition, our proposed method can significantly reduce the storage(memory) space and execution time, compared to the conventional ensemble deep networks developed for face recognition.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2005.11a
• /
• pp.199-202
• /
• 2005

TV OSD(On Screen Display) 메뉴 자동검증 시스템에서 다국적 언어의 문자 인식은 표준패턴의 구조적 분석이 쉽지 않을 뿐만 아니라 학습패턴 집합의 규모와 특징의 수가 증가함으로 인하여 특징추출 및 인식 과정에서 방대한 계산량이 요구된다. 이에 본 연구에서는 학습 데이터에 포함되는 다량의 특징 집합으로부터 인식에 필요한 효과적인 특징을 선별함으로써 패턴 분류기의 효율성을 개선하기 위한 방법론을 고찰한다. 이를 위하여 수정된 형태의 Adaboost 기법을 제안하고 이를 적용한 실험 결과로부터 그 유용성을 고찰한다. 제안된 알고리즘은 초기의 특징 집합을 취약한 성능을 갖는 다수의 분류기(classifier)로서 고려하며, 이로부터 반복학습을 통하여 개선된 분류기를 점진적으로 선별해 나가게 된다. 학습의 원리는 주어진 학습패턴 집합에 기초하여 일종의 교사학습(supervised learning) 방식으로 이루어진다. 각 패턴에 할당된 가중치 값은 각 단계에서 산출되는 분류결과에 따라 적응적으로 수정되어 반복학습이 진행됨에 따라 점차 보완적 성능을 갖는 분류기를 선택할 수 있게 한다. 즉, 주어진 각 학습패턴에 대하여 초기에 균등한 가중치가 부여되며, 반복학습의 각 단계에서 적용되는 분류기의 출력을 분석하여 오분류된 패턴의 가중치 분포를 증가시켜 나간다. 본 연구에서는 실제 응용으로서 OSD 메뉴검증 시스템을 대상으로 제안된 이론을 적용하고 그 타당성을 평가한다.

• Journal of Internet Computing and Services
• /
• v.7 no.1
• /
• pp.59-64
• /
• 2006

Harn al proposed a GKRS(Global Key Recovery System) that combines the functions of the key recovery authorities and the public key certification authorities(CA), Among other features, user dominance(i.e, a user is allowed to select his own public-private key pair and especially a public element for verifying the validity of the public-private key pair)is proposed by [1] for wide acceptance of GKRS. In this paper, we attack the RSA version of GKRS by showing that its user-dominance feature and the corresponding key verification scheme employed by the CA allow for fraud by users against CA. We propose more secure GKPS than original GKPS, The proposed system makes the probability of user fraud negligible small.

• Journal of Information Processing Systems
• /
• v.16 no.1
• /
• pp.201-209
• /
• 2020

The graph data structure is popular because it can intuitively represent real-world knowledge. Graph databases have attracted attention in academia and industry because they can be used to maintain graph data and allow users to mine knowledge. Mining reachability relationships between two nodes in a graph, termed reachability query processing, is an important functionality of graph databases. Online traversals, such as the breadth-first and depth-first search, are inefficient in processing reachability queries when dealing with large-scale graphs. Labeling schemes have been proposed to overcome these disadvantages. The state-of-the-art is the 2-hop labeling scheme: each node has in and out labels containing reachable node IDs as integers. Unfortunately, existing 2-hop labeling schemes generate huge 2-hop label sizes because they only consider local features, such as degrees. In this paper, we propose a more efficient 2-hop label size reduction approach. We consider the topological sort index, which is a global feature. A linear combination is suggested for utilizing both local and global features. We conduct experiments over real-world and synthetic directed acyclic graph datasets and show that the proposed approach generates smaller labels than existing approaches.

• International Journal of Computer Science & Network Security
• /
• v.22 no.9
• /
• pp.403-413
• /
• 2022

Classification and analysis are improved factors for the realtime automation system. In the field of agriculture, the cultivation of different paddy crop depends on the atmosphere and the soil nature. We need to analyze the moisture level in the area to predict the type of paddy that can be cultivated. For this process, Ensemble Modulation Pattern system and Block Probability Neural Network based classification models are used to analyze the moisture and temperature of land area. The dataset consists of the collections of moisture and temperature at various data samples for a land. The Ensemble Modulation Pattern based feature analysis method, the extract of the moisture and temperature in various day patterns are analyzed and framed as the pattern for given dataset. Then from that, an improved neural network architecture based on the block probability analysis are used to classify the data pattern to predict the class of paddy crop according to the features of dataset. From that classification result, the measurement of data represents the type of paddy according to the weather condition and other features. This type of classification model assists where to plant the crop and also prevents the damage to crop due to the excess of water or excess of temperature. The result analysis presents the comparison result of proposed work with the other state-of-art methods of data classification.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.12
• /
• pp.4326-4344
• /
• 2021

The SinGAN is one of generative adversarial networks that can be trained on a single nature image. It has poor ability to learn more global features from nature image, and losses much local detail information when it generates arbitrary size image sample. To solve the problem, a non-linear function is firstly proposed to control downsampling ratio that is ratio between the size of current image and the size of next downsampled image, to increase the ratio with increase of the number of downsampling. This makes the low-resolution images obtained by downsampling have higher proportion in all downsampled images. The low-resolution images usually contain much global information. Therefore, it can help the model to learn more global feature information from downsampled images. Secondly, the attention mechanism is introduced to the generative network to increase the weight of effective image information. This can make the network learn more local details. Besides, in order to make the output image more natural, the TVLoss function is introduced to the loss function of SinGAN, to reduce the difference between adjacent pixels and smear phenomenon for the output image. A large number of experimental results show that our proposed model has better performance than other methods in generating random samples with fixed size and arbitrary size, image harmonization and editing.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.22 no.3
• /
• pp.471-479
• /
• 2018

Recently, the spread of smart TV based Android iOS Set Top box has become common. This paper propose a new approach to control the TV using gestures away from the era of controlling the TV using remote control. In this paper, the AdaBoost algorithm is applied to gesture recognition by using a mono camera. First, we use Camshift-based Body tracking and estimation algorithm based on Gaussian background removal for body coordinate extraction. Using global and local feature vectors, we recognized gestures with speed change. By tracking the time interval trajectories of hand and wrist, the AdaBoost algorithm with CART algorithm is used to train and classify gestures. The principal component feature vector with high classification success rate is searched using CART algorithm. As a result, 24 optimal feature vectors were found, which showed lower error rate (3.73%) and higher accuracy rate (95.17%) than the existing algorithm.

• Journal of KIISE:Software and Applications
• /
• v.33 no.4
• /
• pp.402-410
• /
• 2006

This paper presents a method for feature based block motion estimation that uses multi -resolution image sequences to obtain the panoramic images in the continuous ultrasound image sequences. In the conventional block motion estimation method, the centers of motion estimation blocks are set at the predetermined and equally spaced locations. This requires the large blocks to include at least one feature, which inevitably requires long estimation time. In this paper, we propose an adaptive method which locates the center of the motion estimation blocks at the feature points. This make it possible to reduce the block size while keeping the motion estimation accuracy The Harris-Stephen corner detector is used to get the feature points. The comer points tend to group together, which cause the error in the global motion estimation. In order to distribute the feature points as evenly as Possible, the image is firstly divided into regular subregions, and a strongest corner point is selected as a feature in each subregion. The ultrasound Images contain speckle patterns and noise. In order to reduce the noise artifact and reduce the computational time, the proposed method use the multi-resolution image sequences. The first algorithm estimates the motion in the smoothed low resolution image, and the estimated motion is prolongated to the next higher resolution image. By this way the size of search region can be reduced in the higher resolution image. Experiments were performed on three types of ultrasound image sequences. These were shown that the proposed method reduces both the computational time (from 77ms to 44ms) and the displaced frame difference (from 66.02 to 58.08).

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.4
• /
• pp.41-51
• /
• 2023

In this paper, we propose a technique of multi-time window feature extraction for anger detection in gait data. In the previous gait-based emotion recognition methods, the pedestrian's stride, time taken for one stride, walking speed, and forward tilt angles of the neck and thorax are calculated. Then, minimum, mean, and maximum values are calculated for the entire interval to use them as features. However, each feature does not always change uniformly over the entire interval but sometimes changes locally. Therefore, we propose a multi-time window feature extraction technique that can extract both global and local features, from long-term to short-term. In addition, we also propose an ensemble model that consists of multiple classifiers. Each classifier is trained with features extracted from different multi-time windows. To verify the effectiveness of the proposed feature extraction technique and ensemble model, a public three-dimensional gait dataset was used. The simulation results demonstrate that the proposed ensemble model achieves the best performance compared to machine learning models trained with existing feature extraction techniques for four performance evaluation metrics.