• Journal of IKEEE
• /
• v.23 no.4
• /
• pp.1314-1320
• /
• 2019

In this paper, we propose a new method for optimizing bit operations and applying them to DNN(Deep Neural Network) in software environment. As a method for this, we propose a packing function for bitwise optimization and a masking matrix multiplication operation for application to DNN. The packing function converts 32-bit real value to 2-bit quantization value through threshold comparison operation. When this sequence is over, four 32-bit real values are changed to one 8-bit value. The masking matrix multiplication operation consists of a special operation for multiplying the packed weight value with the normal input value. And each operation was then processed in parallel using a GPU accelerator. As a result of this experiment, memory saved about 16 times than 32-bit DNN Model. Nevertheless, the accuracy was within 1%, similar to the 32-bit model.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.35S no.6
• /
• pp.101-108
• /
• 1998

An efficient approach to recognize occluded objects is to detect a number of essential features on the boundary of the unknown shape. The salient points including corner points, tangential points and inflection points are detected by the relation of neighboring pixels of each pixel on the boundaries. Corner points are usually detected in the curvature function and tangential points and inflection points are detected by median filtering the curvature function to avoid the effect of quantization noise as corner points is not sufficient to represent an object with lines and arcs. Then, these salient points are used as features for object matching. Discrete Hopfield Neural Network is used for object matching. Experimental results show that the matching result using salient points is better than those of using corner points only when an object consists of lines and arcs.

• Journal of KIISE:Software and Applications
• /
• v.36 no.12
• /
• pp.1016-1023
• /
• 2009

A license plate recognition system consists of image processing in which characters and features are extracted, and pattern recognition in which extracted characters are classified. Feature extraction plays an important role in not only the level of data reduction but also performance of recognition. Thus, in this paper, we focused on the recognition of numeral characters especially on the feature extraction of numeral characters which has much effect in the result of plate recognition. We suggest a hybrid statistical feature model which assures the best dispersion of input data by reassignment of clustering property of input data. And we verify the effectiveness of suggested model using multi-layer perceptron and learning vector quantization neural networks. The results show that the proposed feature extraction method preserves the information of a license plate well and also is robust and effective for even noisy and external environment.

• The Korean Journal of Applied Statistics
• /
• v.34 no.6
• /
• pp.979-993
• /
• 2021

Acoustic scene classification (ASC) categorizes an audio file based on the environment in which it has been recorded. This has long been studied in the detection and classification of acoustic scenes and events (DCASE). In this study, we considered the problem that ASC faces in real-world applications that the model used should have low-complexity. We compared several models that apply light-weight techniques. First, a base CNN model was proposed using log mel-spectrogram, deltas, and delta-deltas features. Second, depthwise separable convolution, linear bottleneck inverted residual block was applied to the convolutional layer, and Quantization was applied to the models to develop a low-complexity model. The model considering low-complexity was similar or slightly inferior to the performance of the base model, but the model size was significantly reduced from 503 KB to 42.76 KB.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.49 no.7
• /
• pp.389-398
• /
• 2000

The popular methods to check the identity of individuals include passwords and ID cards. These conventional method for user identification and authentication are not altogether reliable because they can be stolen and forgotten. As an alternative of the existing methods, biometric technology has been paid much attention for the last few decades. In this paper, we propose an efficient system for recognizing the identity of a living person by analyzing iris patterns which have a high level of stability and distinctiveness than other biometric measurements. The proposed system is based on wavelet transform and a competitive neural network with the improved mechanisms. After preprocessing the iris data acquired through a CCD camera, feature vectors are extracted by using Haar wavelet transform. LVQ(Learning Vector Quantization) is exploited to classify these feature vectors. We improve the overall performance of the proposed system by optimizing the size of feature vectors and by introducing an efficient initialization of the weight vectors and a new method for determining the winner in order to increase the recognition accuracy of LVQ. From the experiments, we confirmed that the proposed system has a great potential of being applied to real applications in an efficient and effective way.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.10 no.5
• /
• pp.466-477
• /
• 2000

For developing a CMAC-based adaptive critic learning system to control the cart-pole system, various papers including neural network based learning control schemes as well as an adaptive critic learning algorithm with Adaptive Search Element are reviewed and the adaptive critic learning algorithm for the ASE is integrated into a CMAC controller. Also, quantization problems involved in integrating CMAC into ASE system are studied. By comparing the learning speed of the CMAC system with that of the ASE system and by considering the learning genemlization of the CMAC system with the adaptive critic learning, the applicability of the adaptive critic learning algorithm to CMAC is discussed.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.16 no.6
• /
• pp.545-555
• /
• 1991

In this paper, We discuss optical implementation of new optical adaptive patern recognition system based on single layer perception with learning capability and associative memory model having error corrective capability. The single layer perceptron is optically implemented by using 2 D LCTV spatial light modulators through the nonlinear quantization and polarization encoding methods, and 2 D hopfield associative memory is also implemented by using multifocus holographic lens. From some experimental results on classfication of Arabic numbers into even & edd numbers, it is shown that the proposed system can classify the patterns to the right classes correctly even for the partial and erronenous input patterns. Accordingly, the proposed optical adaptive pattern recognition system can be suggested for practical application in the fields of image processing and pattern recognition.

• Journal of Broadcast Engineering
• /
• v.28 no.1
• /
• pp.141-144
• /
• 2023

Recently, a joint channel estimation, channel quantization, feedback, and precoding system based on deep-neural network (DNN) was proposed. The corresponding system achieved a joint optimization based on deep learning such that it achieved a higher sum rate than the existing codebook-based precoding systems. However, this DNN-based procoding system is not directly applicable for the environments with many users such that a specific user selection can potentially increase the sum rate of the system. Thus, in this letter, we study an appropriate user selection method suitable for DNN-based precoding.

• The Transactions of the Korea Information Processing Society
• /
• v.4 no.1
• /
• pp.149-160
• /
• 1997

This paper is concerned with the problem of speaker-adaptive speech synthes is method using a mapped codebook designed by fuzzy mapping on FLVQ (Fuzzy Learning Vector Quantization). The FLVQ is used to design both input and reference speaker's codebook. This algorithm is incorporated fuzzy membership function into the LVQ(learning vector quantization) networks. Unlike the LVQ algorithm, this algorithm minimizes the network output errors which are the differences of clas s membership target and actual membership values, and results to minimize the distances between training patterns and competing neurons. Speaker Adaptation in speech synthesis is performed as follow;input speaker's codebook is mapped a reference speaker's codebook in fuzzy concepts. The Fuzzy VQ mapping replaces a codevector preserving its fuzzy membership function. The codevector correspondence histogram is obtained by accumulating the vector correspondence along the DTW optimal path. We use the Fuzzy VQ mapping to design a mapped codebook. The mapped codebook is defined as a linear combination of reference speaker's vectors using each fuzzy histogram as a weighting function with membership values. In adaptive-speech synthesis stage, input speech is fuzzy vector-quantized by the mapped codcbook, and then FCM arithmetic is used to synthesize speech adapted to input speaker. The speaker adaption experiments are carried out using speech of males in their thirties as input speaker's speech, and a female in her twenties as reference speaker's speech. Speeches used in experiments are sentences /anyoung hasim nika/ and /good morning/. As a results of experiments, we obtained a synthesized speech adapted to input speaker.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.38 no.2
• /
• pp.205-210
• /
• 2014

Recently, pattern recognition methods have been widely used by researchers for fault diagnoses of mechanical systems. A pattern recognition method determines the soundness of a mechanical system by detecting variations in the system's vibration characteristics. Hidden Markov models (HMMs) and artificial neural networks (ANNs) have recently been used as pattern recognition methods in various fields. In this study, a HMM-ANN hybrid method for the fault diagnosis of a mechanical system is introduced, and a rotating wind turbine blade with a crack is selected for fault diagnosis. The existence, location, and depth of said crack are identified in this research. For improving the diagnostic accuracy of the method in spite of the presence of noise, a moment with a few specific frequencies is applied to the structure.