• Journal of KIISE:Software and Applications
• /
• v.26 no.1
• /
• pp.93-93
• /
• 1999

Researches on digit recognition have been conducted actively for a long time because the classes to recognize are much fewer than other character sets and because it is very likely thatthe digit recognition can be applied to many problems in real world, The recent studies on designingrecognition system with high performance are in progress with two different aspects. One is toconstruct a recognizer using several features at the same time, and the other is to use severalrecognizers. In this paper, we propose a multistage combination method to recognize the unconstrainedhandwritten numerals. The method is a two-stage combination method which uses multiplecombination methods at the same time unlike the existing methods with only one combination method.The recognizers are first combined by several combination methods of different classes simultaneously,and then the results of them are combined by another combination method to generate a final result.Five recognizers and eight combination methods are used in the proposed system. The experimentalresults showed that the recognition rates on CENPARMI and CEDAR data were 97.75% and 98.6%,respectively and the recognition performance could be improved as the process passed through stages,We could get the best performance by combining the combination methods of different classes, whichmeans there are a complementary relation among them, The proposed method can be considered asan extended version of the existing combination methods.

• International Journal of Computer Science & Network Security
• /
• v.22 no.3
• /
• pp.61-66
• /
• 2022

Object classification is one of the main fields in neural networks and has attracted the interest of many researchers. Although there have been vast advancements in this area, still there are many challenges that are faced even in the current era due to its inefficiency in handling large data, linguistic and dimensional complexities. Powerful hardware and software approaches in Neural Networks such as Deep Neural Networks present efficient mechanisms and contribute a lot to the field of object recognition as well as to handle time series classification. Due to the high rate of accuracy in terms of prediction rate, a neural network is often preferred in applications that require identification, segmentation, and detection based on features. Neural networks self-learning ability has revolutionized computing power and has its application in numerous fields such as powering unmanned self-driving vehicles, speech recognition, etc. In this paper, the experiment is conducted to implement a neural approach to identify numbers in different formats without human intervention. Measures are taken to improve the efficiency of the machines to classify and identify numbers. Experimental results show the importance of having training sets to achieve better recognition accuracy.

• Journal of the Korean Society for information Management
• /
• v.18 no.2
• /
• pp.105-124
• /
• 2001

The purpose of this study is to develop a Korean word analysis system, which can improve performance of IRS, based on various methods of word analysis. In this study we focused on maximizing the speed of Korean word analysis, modulizing each functional system and analyzing Korean morpheme precisely. The system, developed in this study, implemented optimal algorithm to increase the speed of word analysis and to verify speed and performance of each subsystem. In addition, the numeral analysis processing was achieved to reduce a system burden by avoiding recursive analysis of compound nouns, based on numeral pattern recognition.

• Journal of KIISE:Software and Applications
• /
• v.34 no.9
• /
• pp.857-868
• /
• 2007

The good classifier ensemble should have a high complementarity among classifiers in order to produce a high recognition rate and its size is small in order to be efficient. This paper proposes a classifier ensemble selection algorithm with coarse-to-fine stages. for the algorithm to be successful, the original classifier pool should be sufficiently diverse. This paper produces a large classifier pool by combining several different classification algorithms and lots of feature subsets. The aim of the coarse selection is to reduce the size of classifier pool with little sacrifice of recognition performance. The fine selection finds near-optimal ensemble using genetic algorithms. A hybrid genetic algorithm with improved searching capability is also proposed. The experimentation uses the worldwide handwritten numeral databases. The results showed that the proposed algorithm is superior to the conventional ones.

• Proceedings of the IEEK Conference
• /
• 1999.06a
• /
• pp.511-515
• /
• 1999

The main factor decreasing speech recognition rate is the surrounding noise. To lower the noise effect, we generally used the filter bank at preprocessing stage. But, in this paper, we tried to recognize the 10 numeral numbers using 2-D LPC to extract image feature. At first, we obtained the result of speech-only recognition using 13th-order LPC coefficients and then, for distorted speech recognition results of ‘0’, ‘4’, ‘5’, ‘6’ and 9’, we added image parameters such as 12th-order 2-D LPC coefficients. At each frame, we extracted the 2-D LPC coefficients, and simulated recognizer with two parameters such as speech and image. Finally, for the numbers, such as ‘4’and ‘9’, the better results were obtained.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.2
• /
• pp.159-167
• /
• 1996

In this paper, a new competitive learning algorithm called dynamic competitive learning (DCL) is presented. DCL is a supervised learning mehtod that dynamically generates output neuraons and nitializes weight vectors from training patterns. It introduces a new parameter called LOG (limit of garde) to decide whether or not an output neuron is created. In other words, if there exist some neurons in the province of LOG that classify the input vector correctly, then DCL adjusts the weight vector for the neuraon which has the minimum grade. Otherwise, it produces a new output neuron using the given input vector. It is largely learning is not limited only to the winner and the output neurons are dynamically generated int he trining process. In addition, the proposed algorithm has a small number of parameters. Which are easy to be determined and applied to the real problems. Experimental results for patterns recognition of remote sensing data and handwritten numeral data indicate the superiority of dCL in comparison to the conventional competitive learning methods.

• Annual Conference on Human and Language Technology
• /
• 2000.10d
• /
• pp.63-67
• /
• 2000

In order to develop a character recognition system, it is an essential preceding work that gathers an image data of the standard. On this purpose a data of the digitized images of a handwritten characters was collected. The types of a gathered image data are Korean character, Chiness character, Numeral, English character, Special character, and so on. This paper deals with a handwritten character image data base, and the image data base different from the general storage structure of a lame capacity multimedia was designed and builded.

• Journal of Electrical Engineering and information Science
• /
• v.1 no.1
• /
• pp.108-117
• /
• 1996

In this paper, we propose a new type of recurrent neural network architecture in which each output unit is connected with itself and fully-connected with other output units and all hidden units. The proposed recurrent network differs from Jordan's and Elman's recurrent networks in view of functions and architectures because it was originally extended from the multilayer feedforward neural network for improving the discrimination and generalization power. We also prove the convergence property of learning algorithm of the proposed recurrent neural network and analyze the performance of the proposed recurrent neural network by performing recognition experiments with the totally unconstrained handwritten numeral database of Concordia University of Canada. Experimental results confirmed that the proposed recurrent neural network improves the discrimination and generalization power in recognizing spatial patterns.

• 한국HCI학회:학술대회논문집
• /
• 2009.02a
• /
• pp.94-98
• /
• 2009

A video caption recognition system has been developed for broadcasting sport events such as major league baseball. The purpose of the system is to translate the information expressed in English units such as miles per hour (MPH) to the international system of units (SI) such as km/h. The system detects the ball speed displayed in the video and recognizes the numerals. The ball speed is then converted to km/h and displayed by the following character generator (CG) system. Although neural-network based methods are widely used for character and numeral recognition, we use template matching to avoid the training process required before the broadcasting. With the proposed template matching method, the operator can cope with the situation when the caption’s appearance changed without any notification. Templates are configured by the operator with a captured screenshot of the first pitch with ball speed. Templates are updated with following correct recognition results. The accuracy of the recognition module is over 97%, which is still not enough for live broadcasting. When the recognition confidence is low, the system asks the operator for the correct recognition result. The operator chooses the right one using hot keys.

• The Transactions of the Korea Information Processing Society
• /
• v.3 no.4
• /
• pp.915-925
• /
• 1996

In this paper, in order to find effective features which can handle variations in off-line handwritten numerals, we performed a comparative study on various sets of features. Results of experimental performance comparison shows that 4- directional features using contours and features which combined cross distance, cross, mesh and projection features are very effective for off-line handwritten numerals recognition in terms of recognition rates and recognition time. And in order to surmount limitation of recognition rate by a single neural network. we proposed a modularized neural network using majority voting and reliability factor with complex feature that mix effective features together. In order to verify the performance of the proposed method, the handwritten numeral databases of Concordia University of Canada and Dong-A University of Korea are used in the experiments. With the database of Concordia University, the recognition rate of 97.1%, the rejection rate of 1.5%, the error rate of 1.4% and the reliability of 98.5% are obtained ; and with the database of Dong-A University, there cognition rate of 98%, the rejection rate of 1.2%, the error rate of 0.8%, the reliability o99.1% are obtained.