• The Transactions of the Korea Information Processing Society
• /
• v.4 no.5
• /
• pp.1277-1288
• /
• 1997

In this paper,the extraciton method for individual and the recognition method for the printed dociments are discussed. In preprocessing is a technique to extract characters that are difficult to manage such as touching characters or overlapped chracters.Genrally in the existing segmentation methods,projection and edge detection are applied.However,in this paper an indvidual character is extracted by using connected pixel with one projection after the string extraction The maximum Blok Methld(MBM)is used for the recognition.The MBM is a method to enlarge the block to the last point the pixel that was found during projection. The maximum blocks are skeletonxied after the division into straight line block and oblique line block.Especially,in the recognition of chinese chracters compared to the existing method it showed improved recognition rate.

• The Transactions of the Korea Information Processing Society
• /
• v.5 no.3
• /
• pp.671-680
• /
• 1998

Network services, such as teleconferencing, remote diagnostics and education, and CSCW require multicasting. Multicast routing methods can be divided into two categories. One is the shortest path tree method and the other is the minimal Steiner tree method. The latter has an advantage over the former in that only one Steiner tree is needed for a group. However, finding a minimal Steiner tree is an NP-complete problem and it is necessary to find an efficient heuristic algorithm. In this paper, we present an evolutionary optimization method for finding minimal Steiner trees without sacrificing too much computational efforts. In particular, we describe a tree-based genetic encoding scheme which is in sharp constast with binary string representations usually adopted in convetional genetic algorithms. Experiments have been performed to show that the presented method can find optimal Steiner trees for given vetwork configurations. Comparitivie studies have shown that the evolutionary method finds on average a better solution than other conventional heustric algorithms.

• Journal of KIISE:Software and Applications
• /
• v.34 no.11
• /
• pp.991-1000
• /
• 2007

A new robust approach to detect vehicles in a single frame of traffic scenes is presented. The method is based on the multi-level shadow classification, which has been shown to have the capability of extracting correct shadow shapes regardless of the operating conditions. The rationale of this classification is supported by the fact that shadow regions underneath vehicles usually exhibit darker gray level regardless of the vehicle brightness and illuminating conditions. Classified shadows provide string clues on the presence of vehicles. Unlike other schemes, neither background nor temporal information is utilized; thereby the performance is robust to the abrupt change of weather and the traffic congestion. By a simple evidential reasoning, the shadow evidences are combined with bright evidences to locate correct position of vehicles. Experimental results show the missing rate ranges form 0.9% to 7.2%, while the false alarm rate is below 4% for six traffic scenes sets under different operating conditions. The processing speed for more than 70 frames per second could be obtained for nominal image size, which makes the real-time implementation of measuring the traffic parameters possible.

• Journal of KIISE:Databases
• /
• v.33 no.2
• /
• pp.163-174
• /
• 2006

In distributed wireless sensor network, it is difficult to transmit and analyze the entire stream data depending on limited networks, power and processor. Therefore it is suitable to use alternative stream data processing after classifying the continuous stream data. We propose a classification framework for continuous multivariate stream data. The proposed approach works in two steps. In the preprocessing step, it takes input as a sliding window of multivariate stream data and discretizes the data in the window into a string of symbols that characterize the signal changes. In the classification step, it uses a standard text classification algorithm to classify the discretized data in the window. We evaluated both supervised and unsupervised classification algorithms. For supervised, we tested Bayesian classifier and SVM, and for unsupervised, we tested Jaccard, TFIDF Jaro and Jaro Winkler. In our experiments, SVM and TFIDF outperformed other classification methods. In particular, we observed that classification accuracy is improved when the correlation of attributes is also considered along with the n-gram tokens of symbols.

• The Transactions of the Korea Information Processing Society
• /
• v.4 no.10
• /
• pp.2596-2604
• /
• 1997

There are two kinds of backends in ACK:code generator(full-fledged backend) and code expander(fast backend). Code generators generate target code using string pattern matching and code expanders generate target code using macro expansion. ACK translates EM to SPARC code using code expander. The corresponding SPARC code sequences for a EM code are generated and then push-pop optimization is performed. But, there is the problem of maintaining hybrid stack. And code expander is not considered to passes parameters of a procedure call through register windows. The purpose of this paper is to improve SPARC code quality. We suggest a method of SPARC cod generation using EM tree. Our method is divided into two phases:EM tree building phase and code expansion phase. The EM tree building phase creates the EM tree and code expansion phase translates it into SPARC code. EM tree is designed to pass parameters of a procedure call through register windows. To remove hybrid stack, we extract an additional information from EM code. We improved many disadvantages that arise from code expander in ACK.

• Science of Emotion and Sensibility
• /
• v.15 no.4
• /
• pp.503-516
• /
• 2012

According to the researches of robot and software agent until now, computational emotion model is dependent on system, so it is hard task that emotion models is separated from existing systems and then recycled into new systems. Therefore, I introduce the Engine of computational Emotion model (shall hereafter appear as EE) to integrate with any robots or agents. This is the engine, ie a software for independent form from inputs and outputs, so the EE is Emotion Generation to control only generation and processing of emotions without both phases of Inputs(Perception) and Outputs(Expression). The EE can be interfaced with any inputs and outputs, and produce emotions from not only emotion itself but also personality and emotions of person. In addition, the EE can be existed in any robot or agent by a kind of software library, or be used as a separate system to communicate. In EE, emotions is the Primary Emotions, ie Joy, Surprise, Disgust, Fear, Sadness, and Anger. It is vector that consist of string and coefficient about emotion, and EE receives this vectors from input interface and then sends its to output interface. In EE, each emotions are connected to lists of emotional experiences, and the lists consisted of string and coefficient of each emotional experiences are used to generate and process emotional states. The emotional experiences are consisted of emotion vocabulary understanding various emotional experiences of human. This study EE is available to use to make interaction products to response the appropriate reaction of human emotions. The significance of the study is on development of a system to induce that person feel that product has your sympathy. Therefore, the EE can help give an efficient service of emotional sympathy to products of HRI, HCI area.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.15 no.1
• /
• pp.56-61
• /
• 2022

In this paper, we classify and present AI algorithms and natural language processing methods used in chatbots. A framework that can be used to implement a chatbot is also described. A chatbot is a system with a structure that interprets the input string by constructing the user interface in a conversational manner and selects an appropriate answer to the input string from the learned data and outputs it. However, training is required to generate an appropriate set of answers to a question and hardware with considerable computational power is required. Therefore, there is a limit to the practice of not only developing companies but also students learning AI development. Currently, chatbots are replacing the existing traditional tasks, and a practice course to understand and implement the system is required. RNN and Char-CNN are used to increase the accuracy of answering questions by learning unstructured data by applying technologies such as deep learning beyond the level of responding only to standardized data. In order to implement a chatbot, it is necessary to understand such a theory. In addition, the students presented examples of implementation of the entire system by utilizing the methods that can be used for coding education and the platform where existing developers and students can implement chatbots.

• Journal of Intelligence and Information Systems
• /
• v.26 no.2
• /
• pp.1-25
• /
• 2020

In this paper, we suggest an application system architecture which provides accurate, fast and efficient automatic gasometer reading function. The system captures gasometer image using mobile device camera, transmits the image to a cloud server on top of private LTE network, and analyzes the image to extract character information of device ID and gas usage amount by selective optical character recognition based on deep learning technology. In general, there are many types of character in an image and optical character recognition technology extracts all character information in an image. But some applications need to ignore non-of-interest types of character and only have to focus on some specific types of characters. For an example of the application, automatic gasometer reading system only need to extract device ID and gas usage amount character information from gasometer images to send bill to users. Non-of-interest character strings, such as device type, manufacturer, manufacturing date, specification and etc., are not valuable information to the application. Thus, the application have to analyze point of interest region and specific types of characters to extract valuable information only. We adopted CNN (Convolutional Neural Network) based object detection and CRNN (Convolutional Recurrent Neural Network) technology for selective optical character recognition which only analyze point of interest region for selective character information extraction. We build up 3 neural networks for the application system. The first is a convolutional neural network which detects point of interest region of gas usage amount and device ID information character strings, the second is another convolutional neural network which transforms spatial information of point of interest region to spatial sequential feature vectors, and the third is bi-directional long short term memory network which converts spatial sequential information to character strings using time-series analysis mapping from feature vectors to character strings. In this research, point of interest character strings are device ID and gas usage amount. Device ID consists of 12 arabic character strings and gas usage amount consists of 4 ~ 5 arabic character strings. All system components are implemented in Amazon Web Service Cloud with Intel Zeon E5-2686 v4 CPU and NVidia TESLA V100 GPU. The system architecture adopts master-lave processing structure for efficient and fast parallel processing coping with about 700,000 requests per day. Mobile device captures gasometer image and transmits to master process in AWS cloud. Master process runs on Intel Zeon CPU and pushes reading request from mobile device to an input queue with FIFO (First In First Out) structure. Slave process consists of 3 types of deep neural networks which conduct character recognition process and runs on NVidia GPU module. Slave process is always polling the input queue to get recognition request. If there are some requests from master process in the input queue, slave process converts the image in the input queue to device ID character string, gas usage amount character string and position information of the strings, returns the information to output queue, and switch to idle mode to poll the input queue. Master process gets final information form the output queue and delivers the information to the mobile device. We used total 27,120 gasometer images for training, validation and testing of 3 types of deep neural network. 22,985 images were used for training and validation, 4,135 images were used for testing. We randomly splitted 22,985 images with 8:2 ratio for training and validation respectively for each training epoch. 4,135 test image were categorized into 5 types (Normal, noise, reflex, scale and slant). Normal data is clean image data, noise means image with noise signal, relfex means image with light reflection in gasometer region, scale means images with small object size due to long-distance capturing and slant means images which is not horizontally flat. Final character string recognition accuracies for device ID and gas usage amount of normal data are 0.960 and 0.864 respectively.

• 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.

• The Journal of the Korea Contents Association
• /
• v.8 no.1
• /
• pp.68-74
• /
• 2008

Application areas of Knowledge Discovery in Database(KDD) have been expanded to many R&D management processes including technology trends analysis, forecasting and evaluation etc. Established research field such as informetrics (or scientometrics) has utilized techniques or methods of KDD. Various systems have been developed to support works of analyzing large-scale R&D related databases such as patent DB or bibliographic DB by a few researchers or institutions. But extant systems have some problems for korean users to use. Their prices is not moderate, korean language processing is impossible, and user's demands not reflected. To solve these problems, Korea Institute of Science and Technology Information(KISTI) developed stand-alone type information analysis system named as KnowledgeMatrix. KnowledgeMatrix system offer various functions to analyze retrieved data set from databases. KnowledgeMatrix's main operation unit is composed of user-defined lists and matrix generation, cluster analysis, visualization, data pre-processing. Matrix generation unit help extract information items which will be analyzed, and calculate occurrence, co-occurrence, proximity of the items. Cluster analysis unit enable matrix data to be clustered by hierarchical or non-hierarchical clustering methods and present tree-type structure of clustered data. Visualization unit offer various methods such as chart, FDP, strategic diagram and PFNet. Data pre-processing unit consists of data import editor, string editor, thesaurus editor, grouping method, field-refining methods and sub-dataset generation methods. KnowledgeMatrix show better performances and offer more various functions than extant systems.