• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.1287-1292
• /
• 2005

This paper presents an effective pattern classification model by designing an artificial neural network based pattern classifiers for face recognition. First, a RGB image inputted from a frame grabber is converted into a HSV image which is similar to the human beings' vision system. Then, the coarse facial region is extracted using the hue(H) and saturation(S) components except intensity(V) component which is sensitive to the environmental illumination. Next, the fine facial region extraction process is performed by matching with the edge and gray based templates. To make a light-invariant and qualified facial image, histogram equalization and intensity compensation processing using illumination plane are performed. The finally extracted and enhanced facial images are used for training the pattern classification models. The proposed H-ART2 model which has the hierarchical ART2 layers and F-LVQ model which is optimized by fuzzy membership make it possible to classify facial patterns by optimizing relations of clusters and searching clustered reference patterns effectively. Experimental results show that the proposed face recognition system is as good as the SVM model which is famous for face recognition field in recognition rate and even better in classification speed. Moreover high recognition rate could be acquired by combining the proposed neural classification models.

• International Journal of Computer Science & Network Security
• /
• v.21 no.9
• /
• pp.281-291
• /
• 2021

The Islamic jurisprudential legal system represents an essential component of the Islamic religion, that governs many aspects of Muslims' daily lives. This creates many questions that require interpretations by qualified specialists, or Muftis according to the main sources of legislation in Islam. The Islamic jurisprudence is usually classified into branches, according to which the questions can be categorized and classified. Such categorization has many applications in automated question-answering systems, and in manual systems in routing the questions to a specialized Mufti to answer specific topics. In this work we tackle the problem of automatic categorisation of Islamic jurisprudential legal questions using deep learning techniques. In this paper, we build a hierarchical deep learning model that first extracts the question text features at two levels: word and sentence representation, followed by a text classifier that acts upon the question representation. To evaluate our model, we build and release the largest publicly available dataset of Islamic questions and answers, along with their topics, for 52 topic categories. We evaluate different state-of-the art deep learning models, both for word and sentence embeddings, comparing recurrent and transformer-based techniques, and performing extensive ablation studies to show the effect of each model choice. Our hierarchical model is based on pre-trained models, taking advantage of the recent advancement of transfer learning techniques, focused on Arabic language.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.34-36
• /
• 2022

In the existing text sentiment analysis models, the entire text is usually directly modeled as a whole, and the hierarchical relationship between text contents is less considered. However, in the practice of sentiment analysis, many texts are mixed with multiple emotions. If the semantic modeling of the whole is directly performed, it may increase the difficulty of the sentiment analysis model to judge the sentiment, making the model difficult to apply to the classification of mixed-sentiment sentences. Therefore, this paper proposes a sentiment analysis model BHGCN that considers the text hierarchy. In this model, the output of hidden states of each layer of BERT is used as a node, and a directed connection is made between the upper and lower layers to construct a graph network with a semantic hierarchy. The model not only pays attention to layer-by-layer semantics, but also pays attention to hierarchical relationships. Suitable for handling mixed sentiment classification tasks. The comparative experimental results show that the BHGCN model exhibits obvious competitive advantages.

• The KIPS Transactions:PartB
• /
• v.10B no.4
• /
• pp.443-450
• /
• 2003

In this paper, we proposed the novel hierarchical algorithm for the recognition of English calling cards that processes multiresolution images of calling cards hierarchically to extract individual characters and recognizes the extracted characters by using the enhanced neural network method. The hierarchical recognition algorithm generates multiresolution images of calling cards, and each processing step in the algorithm selects and processes the image with suitable resolution for lower processing overhead and improved output. That is, first, the image of 1/3 times resolution, to which the horizontal smearing method is applied, is used to extract the areas including only characters from the calling card image, and next, by applying the vertical smearing and the contour tracking masking, the image of a half time resolution is used to extract individual characters from the character string areas. Lastly, the original image is used in the recognition step, because the image includes the morphological information of characters accurately. And for the recognition of characters with diverse font types and various sizes, the enhanced RBF network that improves the middle layer based on the ART1 was proposed and applied. The results of experiments on a large number of calling card images showed that the proposed algorithm is greatly improved in the performance of character extraction and recognition compared with the traditional recognition algorithms.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.17 no.6
• /
• pp.792-798
• /
• 2007

In this paper, we propose a hierarchical fuzzy-rough classification method based on statistical information for maximizing the performance of pattern classification and reducing the number of rules without learning approaches such as neural network, genetic algorithm. In the proposed method, statistical information is used for extracting the partition intervals of antecedent fuzzy sets at each layer on hierarchical fuzzy-rough classification systems and rough sets are used for minimizing the number of fuzzy if-then rules which are associated with the partition intervals extracted by statistical information. To show the effectiveness of the proposed method, we compared the classification results(e.g. the classification accuracy and the number of rules) of the proposed with those of the conventional methods on the Fisher's IRIS data. From the experimental results, we can confirm the fact that the proposed method considers only statistical information of the given data is similar to the classification performance of the conventional methods.

• Journal of the Institute of Convergence Signal Processing
• /
• v.20 no.4
• /
• pp.205-211
• /
• 2019

Steganalysis is a technique that aims to detect and recover data hidden by steganography. Steganalytic methods detect hidden data by analyzing visual and statistical distortions caused during data embedding. However, for recovering the hidden data, they need to know which steganographic methods the hidden data has been embedded by. Therefore, we propose a hierarchical convolutional neural network (CNN) structure that identifies a steganographic method applied to an input image through multi-level classification. We trained four base CNNs (each is a binary classifier that determines whether or not a steganographic method has been applied to an input image or which of two different steganographic methods has been applied to an input image) and connected them hierarchically. Experimental results demonstrate that the proposed hierarchical CNN structure can identify four different steganographic methods (LSB, PVD, WOW, and UNIWARD) with an accuracy of 79%.

• Journal of Digital Contents Society
• /
• v.19 no.2
• /
• pp.415-423
• /
• 2018

In recent years, Computer-based learning, such as machine learning and deep learning in the computer field, is attracting attention. They start learning from the lowest level and propagate the result to the highest level to calculate the final result. Research literature has shown that systematic learning and growth can yield good results. However, systematic models based on systematic models are hard to find, compared to various and extensive research attempts. To this end, this paper proposes the first TNT(Transitive Nested Triangle)model, which is a growth and fusion model that can be used in various aspects. This model can be said to be a recursive model in which each function formed through geometric forms an organic hierarchical relationship, and the result is used again as they grow and converge to the top. That is, it is an analytical method called 'Horizontal Sibling Merges and Upward Convergence'. This model is applicable to various aspects. In this study, we focus on explaining the TNT model.

• The Journal of the Korea Contents Association
• /
• v.10 no.1
• /
• pp.10-18
• /
• 2010

Due to a large number of classes in Hangul character recognition, it is usual to use the six-type preclassification stage. After the preclassification, the first consonent, vowel, and last consonent can be classified separately. Though each of three components has a few of classes, classification errors occurs often due to shape similarity such as 'ㅔ' and 'ㅖ'. So this paper proposes a hierarchical recognition method which adopts multi-stage tree structures for each of 6-types. In addition, to reduce the interference among three components, the method uses the recognition results of first consonents and vowel as features of vowel classifier. The recognition accuracy for the test set of PHD08 database was 98.96%.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.34S no.4
• /
• pp.33-41
• /
• 1997

Korean unvoiced phonemes consist of nonstationary parts comparing that the vowels and nasal consonants consist of quasi-stationary part. And some phonemes, which have smae point of articulation but differnt manner of articulation, has similar characteristics, so it makes to be hard to distinguish each other. A new method usin gchanges and characteristics of acoustic properties of these phonemes to improve recognition rate are proposed. And because these changes and cahracteristics evidently occur in continuous speech except some unvoiced consonants are articulated as voiced phoneme in case to be used as an midial between voiced phonemes, this method can be applied easily. The features of the frames extracted to represent each phonemes are used asinputs to the hierarchical neural network. And with these results final decision for phoneme recognition is made thorugh post processing which the new method is applied to. Through the experimental recognition results for 9 unvoiced consonants which belong to bilabial, alveolar, and velar phoneme series, 89.4% recognition rate to distinguish in same phoneme series is obtained, and 85.6% recognition rate is obtained in case of including cistinguishing phoneme series.

• Proceedings of the Korean Information Science Society Conference
• /
• 2012.06b
• /
• pp.321-323
• /
• 2012

스마트폰 사용의 보편화와 센서기술의 발달로 이를 응용하는 다양한 연구가 진행되고 있다. 특히 가속도, GPS, 조도, 방향센서 등의 센서들이 스마트폰에 부착되어 출시되고 있어서, 이를 이용한 상황인지, 행동인식 등의 관련 연구들이 활발하다. 하지만 다양한 클래스를 분류하면서 높은 인식률을 유지하는 것은 어려운 문제이다. 본 논문에서는 인식률 향상을 위해 계층적 구조의 순환 신경망을 이용하여 제스처를 인식한다. 스마트폰의 가속도 센서를 이용하여 사용자의 제스처 데이터를 수집하고 BLSTM(Bidirectional Long Short-Term Memory) 구조의 순환신경망을 계층적으로 사용하여, 20가지 사용자의 제스처와 비제스처를 분류한다. 약 24,850개의 시퀀스 데이터를 사용하여 실험한 결과, 기존 BLSTM은 평균 89.17%의 인식률을 기록한 반면 계층적 BLSTM은 평균 91.11%의 인식률을 나타내었다.