• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.35S no.5
• /
• pp.75-82
• /
• 1998

Conventional vector quantization (VQ) codebook design methods have several drawbacks such as edge degradation and high computational complexity. In this paper, we first made activity coordinates from the horizonatal and the vertical activity of the input block. Then it is mapped on the 2-dimensional interconnected codebook, and the codebook is designed using kohonen self-organizing map (KSFM) learning algorithm after the search of a codevector that has the minumum distance from the input vector in a small window, centered by the mapped point. As the serch area is restricted within the window, the computational amount is reduced compared with usual VQ. From the resutls of computer simulation, proposed method shows a better perfomance, in the view point of edge reconstruction and PSNR, than previous codebook training methods. And we also obtained a higher PSNR than that of classified vector quantization (CVQ).

• Journal of Korea Multimedia Society
• /
• v.7 no.12
• /
• pp.1737-1744
• /
• 2004

In the compression methods widely used today, the image compression by VQ is the most popular and shows a good data compression ratio. Almost all the methods by VQ use the LBG algorithm that reads the entire image several times and moves code vectors into optimal position in each step. This complexity of algorithm requires considerable amount of time to execute. To overcome this time consuming constraint, we propose an enhanced self-organizing neural network for color images. VQ is an image coding technique that shows high data compression ratio. In this study, we improved the competitive learning method by employing three methods for the generation of codebook. The results demonstrated that compression ratio by the proposed method was improved to a greater degree compared to the SOM in neural networks.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 1999.05a
• /
• pp.361-369
• /
• 1999

In this paper, we propose a new intelligent control system to be adapted for the characteristic of user who use the plant. The proposed intelligent control system is composed of the artificial neural network, the teaming vector quantization network. In order to verify the usefulness of the proposed system, we simulated in using Matlab.

• The Journal of the Acoustical Society of Korea
• /
• v.15 no.6
• /
• pp.5-11
• /
• 1996

This paper is the study for recognizing single vowels of speaker-independent, and we suppose a method of speech recognition using VQ(Vector Quantization)/NN(Neural Network). This method makes a VQ codebook, which is used for obtaining the observation sequence, and then claculates the probability value by comparing each codeword with the data, finally uses these probability values for the input value of the neural network. Korean signle vowels are selected for our recognition experiment, and ten male speakers pronounced eight single vowels ten times. We compare the performance of our method with those of fuzzy VQ/HMM and conventional VQ/NN According to the experiment result, the recognition rate by VQ/NN is 92.3%, by VQ/HMM using fuzzy is 93.8% and by VQ/NN using fuzzy is 95.7%. Therefore, it is shown that recognition rate of speech recognition by fuzzy VQ/NN is better than those of fuzzy VQ/HMM and conventional VQ/HMM because of its excellent learning ability.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.12 no.10
• /
• pp.799-807
• /
• 2002

The vibration signal can give an indication of the condition of rotating machinery, highlighting potential faults such as unbalance, misalignment and bearing defects. The features in the vibration signal provide an important source of information for the faults diagnosis of rotating machinery. When additional training data become available after the initial training is completed, the conventional neural networks (NNs) must be retrained by applying total data including additional training data. This paper proposes the fault diagnostics algorithm using the ART-Kohonen network which does not destroy the initial training and can adapt additional training data that is suitable for the classification of machine condition. The results of the experiments confirm that the proposed algorithm performs better than other NNs as the self-organizing feature maps (SOFM) , learning vector quantization (LYQ) and radial basis function (RBF) NNs with respect to classification quality. The classification success rate for the ART-Kohonen network was 94 o/o and for the SOFM, LYQ and RBF network were 93 %, 93 % and 89 % respectively.

• Journal of Institute of Control, Robotics and Systems
• /
• v.6 no.8
• /
• pp.710-718
• /
• 2000

In this paper we described an approach to automation of visual inspection of solder joint defects of SMC(Surface Mounted Components) on PCBs(Printed Circuit Board) by using neural network and fuzzy rule-based classification method. Inherently the surface of the solder joints is curved tiny and specular reflective it induces difficulty of taking good image of the solder joints. And the shape of the solder joints tends to greatly vary with the soldering condition and the shapes are not identical to each other even though the solder joints belong to a set of the same soldering quality. This problem makes it difficult to classify the solder joints according to their qualities. Neural network and fuzzy rule-based classification method is proposed to effi-ciently make human-like classification criteria of the solder joint shapes. The performance of the proposed approach is tested on numerous samples of commercial computer PCB boards and compared with the results of the human inspector performance and the conventional Kohonen network.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.7
• /
• pp.11-20
• /
• 2024

In this paper, we propose the BAG (Bit-width Aware Generator) and the Intermediate Layer Knowledge Distillation using Channel-wise Attention to reduce the knowledge gap between a quantized network, a full-precision network, and a generator in GDFQ (Generative Data-Free Quantization). Since the generator in GDFQ is only trained by the feedback from the full-precision network, the gap resulting in decreased capability due to low bit-width of the quantized network has no effect on training the generator. To alleviate this problem, BAG is quantized with same bit-width of the quantized network, and it can generate synthetic images, which are effectively used for training the quantized network. Typically, the knowledge gap between the quantized network and the full-precision network is also important. To resolve this, we compute channel-wise attention of outputs of convolutional layers, and minimize the loss function as the distance of them. As the result, the quantized network can learn which channels to focus on more from mimicking the full-precision network. To prove the efficiency of proposed methods, we quantize the network trained on CIFAR-100 with 3 bit-width weights and activations, and train it and the generator with our method. As the result, we achieve 56.14% Top-1 Accuracy and increase 3.4% higher accuracy compared to our baseline AdaDFQ.

• 제어로봇시스템학회:학술대회논문집
• /
• 1997.10a
• /
• pp.1628-1630
• /
• 1997

The purpose of this study is to see the method of the analysis of EEG(Electroencephalography) whcih is a nonlinear system, to quantize human emotion under color stimulation using the analysis of EEG. The result of this study would be used clinical study and development fo image instruments with color. In this study, the method of the analysis of EEG is power spectrum using FFT(Fast Fourier Transform) and the modelling of EEG under color stimulation base on back propagation Neural Networks ond of AI(Artfical Intellignece) skills. First, input layer make a match to relative power which get analyzing s in 4 channels, and output layer make a match to color stimulation which is measured human emotion. Finally, weights of each neurons determine by learing back porpagation Neural Networks.

• Journal of Mechanical Science and Technology
• /
• v.16 no.12
• /
• pp.1633-1642
• /
• 2002

Uninterrupted power supply has become indispensable during the maintenance task of active electric power lines as a result of today's highly information-oriented society and increasing demand of electric utilities. The maintenance task has the risk of electric shock and the danger of falling from high place. Therefore it is necessary to realize an autonomous robot system using electro-hydraulic manipulator because hydraulic manipulators have the advantage of electric insulation. Meanwhile it is relatively difficult to realize autonomous assembly tasks particularly in the case of manipulating flexible objects such as electric lines. In this report, a discrete event control system is introduced for automatic assembly task of electric lines into sleeves as one of the typical task of active electric power lines. In the implementation of a discrete event control system, LVQNN (linear vector quantization neural network) is applied to the insertion task of electric lines to sleeves. In order to apply these proposed control system to the unknown environment, virtual learning data for LVQNN is generated by fuzzy inference. By the experimental results of two types of electric lines and sleeves, these proposed discrete event control and neural network learning algorithm are confirmed very effective to the insertion tasks of electric lines to sleeves as a typical task of active electric power maintenance tasks.

• Journal of Power System Engineering
• /
• v.7 no.2
• /
• pp.29-35
• /
• 2003

The monitoring and diagnostics of the rotating machinery have been received considerable attention for many years. The objectives are to classify the machinery condition and to find out the cause of abnormal condition. This paper describes a classification method of diagnosing the small reciprocating compressor for refrigerators using the artificial neural network and the wavelet transform. In order to extract salient features, the wavelet transform are used from primary noise signals. Since the wavelet transform decomposes raw time-waveform signals into two respective parts in the time space and frequency domain, more and better features can be obtained easier than time-waveform analysis. In the training phase for classification, self-organizing feature map(SOFM) and learning vector quantization(LVQ) are applied, and the accuracies of them ate compared with each other. This paper is focused on the development of an advanced signal classifier to automatize the vibration signal pattern recognition. This method is verified by small reciprocating compressors, for refrigerator and normal and abnormal conditions are classified with high flexibility and reliability.