Geri Dön

Hücresel yapay sınır ağları ile alarm seslerinin sınıflandırılması

Başlık çevirisi mevcut değil.

  1. Tez No: 55572
  2. Yazar: İBRAHİM GENÇ
  3. Danışmanlar: DOÇ.DR. CÜNEYT GÜZELİŞ
  4. Tez Türü: Yüksek Lisans
  5. Konular: Elektrik ve Elektronik Mühendisliği, Electrical and Electronics Engineering
  6. Anahtar Kelimeler: Hücresel sinir ağları, Ses analizi, Yapay sinir ağları, Cellular neural networks, Voice analysis, Artificial neural networks
  7. Yıl: 1996
  8. Dil: Türkçe
  9. Üniversite: İstanbul Teknik Üniversitesi
  10. Enstitü: Fen Bilimleri Enstitüsü
  11. Ana Bilim Dalı: Belirtilmemiş.
  12. Bilim Dalı: Belirtilmemiş.
  13. Sayfa Sayısı: Belirtilmemiş.

Özet

ÖZET Bu tezde, Hücresel Yapay Sinir Ağlan (HYSA)'nın alarm seslerinin sınıflandırılmasında kullanılabilirliği incelenmiş ve bu amaçla dalgacık (“wavelet”) dönüşümü (DD)'ne dayalı bir teknik geliştirilmiştir. Böylece, iki boyutlu ızgara biçimindeki mimarisi ile doğal uygulama alanım görüntü işlemede bulan HYSA'nın ses işareti gibi bir boyutlu işaretlerin işlenmesinde kullanımına yönelik ilk adımlardan biri atılmıştır. Tez çalışmasında önerilen yaklaşım, araba korna sesleri, elektronik ve elektromekanik telefon zili sesleri gibi göreceli olarak basit seslerin sınıflandırılması problemine uygulanmış ve sonuçlar literatürdeki benzer çalışmalar ile karşılaştırılmıştır. Sınıflandırma probleminin çözümünde yapılan işlemler üç ana bölüme ayrılarak incelenebilir. Ön işlem birimi olarak adlandırılan birinci bölümde ses işaretleri görüntüye dönüştürülür. Bu birimin çıkışında, dizinin her bir elemanı bir ses çerçevesine (“frame”)'e karşı düşmek üzere bir görüntü dizisi elde edilir. İkinci birim ise zaman içerisinde gelen görüntü dizisinden kolay sınıflandırılabilir tek bir işaret elde etmek amacıyla kullanılır. Bu birim içerisinde bir HYSA ve basit bir mantık işlemcisi bulunmaktadır. Üçüncü ve son birim ise sınıflandırma problemlerinde sıkça kullanılan çok çıkışlı, tek-katmanlı doğrusal bir algılayıcıdan oluşmaktadır. Elde edilen sonuçlar aynı uygulama üzerine yapılan diğer çalışmalar ile karşılaştırıldığında onların verdiklerine yakın mertebede iyi hata oranlan gözlenmiştir. Bu çalışmadaki yaklaşımın bir çok noktada iyileştirmeye olanak tanıması bir üstünlüğü olarak belirtilebilir.

Özet (Çeviri)

SUMMARY This thesis presents Cellular Neural Networks (CNNs)[l],[2] based classification system for recognizing acoustic alarm signals. The proposed classification system uses wavelet transformation (WT) for transforming a one dimensional (1-D) acoustic signal into a sequence of two dimensional (2-D) signals, so called input images, which will be fed to the CNN. Test samples from three sets of acoustical alarm signals are recognized with an error rate of less than 16%, which shows that CNN can be used as a tool for acoustic signal recognition. The extensive works done in the literature show that CNNs with their 2-D grid architecture are well suited for image processing. Up to now, a very limited work is devoted to the applications of CNNs to process 1-D signals such as acoustic signals. This work extends by introducing a wavelet based technique, the field of CNN applications on the classification of acoustic alarm signals which was initiated by Osuna et. al [3]. The proposed solution in this thesis for the classification of acoustical alarm signals is divided into three parts. The first, a transformation of 1-D into a sequence of images is done by WT together with a grid like correlation process. The produced image sequences serve as the inputs to the CNN. CNN is responsible for the extraction of the relevant information carried by the image sequences, and it also performs a time-to-space mapping via concentrating the information distributed over an image sequence into a single image. Thirdly, a simple classification device based on linear perceptron is used for separating the VIdifferent classes of input signals by sorting the output images of the CNN into the correct classes. (See Figure 1.) The first unit in Figure 1, performs a transformation of the 1-D acoustical signal into image sequences. Every input signal is then characterized by its corresponding image sequence. Image sequences produced by the first unit are applied to the second unit, CNN. This block essentially consists of two parts: i) Analysis and ii) Correlation. Signal 1-D to 2-D Transform CNN Classifier Class Sequence of Images Two Images Figure 1. Processing units for the classification of acoustical signals. To perform analysis, WT is used, since there are many similarities between the WT and hearing. In the modelling of cochlea and auditory processing, WT is used by several researchers[6],[7]. Although continuous wavelet transformation can work in this method, discrete wavelet transform (DWT) which is an application of WT for sampled signals, has been used. The reason is that there are no fully programmable CNN chips available yet, and CNN must be simulated on a digital computer. vnWT, simply, decomposes an arbitrary function into a two-parameter family of elementary wavelets that are obtained by shifts in time variable but also dilations (or compressions) that act both on time and the frequency variables. WT can be defined as follows; WTx(x,a) = 4TJx(t)-hf^dt (1) VH ^ a ' where h(t) is a basic wavelet prototype and called as mother wavelet or analysing wavelet, h ft-x is scaled and translated version of basic wavelet V a j prototype, a is scaling parameter, and only positive dilations a>0 is taken into account, x shifts the wavelet in time. h(t) can be any band-pass function. In particular one can dispense with complex-valued transforms and deal only with real-valued ones. The DWT is defined as DWT{x[n];2j,k2j} = cik = 2>(n)h*(n-k2j) (2) where hj(n-k2J)'s are analysis discrete wavelets. In this equation scale, a, is 2İ and time lo cation, x, is k2J. Consider two filter impulse responses g(n) and h(n). Here h stands for high-pass or discrete wavelet and g stands for low-pass. The wavelets and scaling sequences are obtained iteratively as g1(n) = g(n), h1(n) = h(n), (3a) Vlllgj+1(n) = £gj(k)-g(n-2k), hj+1(n) = 2hj(k)-g(n-2k)3 (3b) (3c) (3b) and (3 c) means that one goes from one octave j to the next (j+1) by applying the interpolation operator fj+1(n) = Zfi(k)'g(n-2k)> (4) which should be thought of as the discrete equivalent to the dilation; f(t)->2“1/2f(t/2) The DWT corresponds to the analysis filter bank. The filters present in the filter bank are precisely g(n), h(n). There are several algorithms to implement DWT [12]-[14]. Direct implementation will be employed in this thesis. (See Figure 2.) LPF -® Decimate! by2 HPF © LPF -©- HPF LPF -d>-> k+1 HPF Figure 2. DWT processing block (dotted rectangle) and DWT structure. IXA pseudo code for implementing DWT can be given as folllows: t=x; for every scale { iflength(t) odd, add a Oto the end oft yl=convolution(t,g); // Lowpass filtering yh=convolution(t,h); // Highpass filtering for all n, n<length(yl)/2 { yl(n)=yl(2n); // subsampling by two yh(n)=yh(2n); // subsampling by two } t=yl; // input for the next stage keep yh as wavelet coefficients for this scale } Figure 3. Correlation grid. The 2-D transformation of the acoustical signal is performed in the following way: the same signal is applied to two identical DWT blocks. In simulations, two DWT blocks are not needed, since same coefficients can be used for two blocks. The first placed horizontally and the second vertically at the topand left edge of a square, respectively, both of them having the lowest scale (highest frequency) at the top left corner. The crossings of all DWT coefficients (see Figure 3.) form the pixels of the 2-D representation which is computed at a rate of r frames per second. CNN is confronted with a series of gray-scale images for every alarm signal. Processing method is very simple. A linear template given in Table 1, called square template is used to obtain one image from image sequence. And also to keep changes on the output of CNN during the period of processing, OR logical operation applied. At the end we have two images to classify. (See Figure 4.) (a) (b) Figure 4. (a) Output image of CNN and (b) OR operated output image for a phone ring. The two images are symmetrical so the information to supply perceptron is nearly halved. Classifier is a one-layer linear perceptron with three output units, one for each class. (See Figure 5.) Each vector defines a point in space which stands for the corresponding images. Similar images are mapped onto points close XIto one another in input space, because similarity means that pixel values, and hence vector components, do not differ too much. On the other hand, if the images are distinct, the corresponding points in space are far away from each other. Thus, different classes are clouds of points in space. horn electromechanic phone ring electronic phone ring Figure 5. One layer perceptron with three output units is used in order to classify images. Every neuron of Figure 5 lays a hyperplane in space in order to separate the points which belong to ”its" class, from other points. An input vector which, e.g. belongs to the class of horns, is correctly detected if the output of the first neuron is positive, and all other neurons produce negative outputs. Some results achieved by other methods are given in Table 2 and Table 3 lists the classification error rates achieved with the processing method mentioned above for three different mother wavelet. As seen in the table, some mother wavelets do not give good results while another one does. xnTable 1. The CNN template used in this thesis. A: l=-0.5 Table 2. Classification error rates of other methods [3],[16],[17],[18]. XlllTable 3. Classification error rates achieved in this thesis. XIV

Benzer Tezler

  1. Tez No

    39613

    Hücresel yapay sinir ağları için iki öğrenme algoritması ve görüntü işleme uygulamaları

    Two learning algorithms for cellular neural networks and their image processing applications

    SİNAN KARAMAHMUT

    Yüksek Lisans

    Türkçe

    Türkçe

    1994

    Elektrik ve Elektronik Mühendisliğiİstanbul Teknik Üniversitesi

    DOÇ.DR. CÜNEYT GÜZELİŞ

  2. Tez No

    46172

    F(x)-x bağlantılı bir kaotik hücresel yapay sinir ağının numerik olarak incelenmesi

    Başlık çevirisi yok

    FATİH KAVASLAR

    Yüksek Lisans

    Türkçe

    Türkçe

    1995

    Elektrik ve Elektronik Mühendisliğiİstanbul Teknik Üniversitesi

    DOÇ.DR. CÜNEYT GÜZELİŞ

  3. Tez No

    46271

    Dinamik algılayıcı öğrenme algoritması ile kenar saptamanın öğrenilmesi

    Learning of edge detection using recurrent perceptron learning algorithm

    FİLİZ YOSMA TAŞKIN

    Yüksek Lisans

    Türkçe

    Türkçe

    1995

    Elektrik ve Elektronik Mühendisliğiİstanbul Teknik Üniversitesi

    DOÇ.DR. CÜNEYT GÜZELİŞ

  4. Tez No

    96287

    Hücresel üretim sistemi tasarımında kullanılan yapay zeka teknikleri ile sezgisel yöntemlerin karşılaştırılması ve uygulamalı analizleri

    Comparison of artificial intelligence techniques and heuristic methods used in cellular manufacturing system design and their applied analysis

    YAKUP KARA

    Yüksek Lisans

    Türkçe

    Türkçe

    2000

    Endüstri ve Endüstri MühendisliğiSelçuk Üniversitesi

    Endüstri Mühendisliği Ana Bilim Dalı

    DOÇ.DR. BİLAL TOKLU

  5. Tez No

    29683

    Yapay sinir ağları ve uygulamaları

    Başlık çevirisi yok

    NURCİHAN ÜNALDI

    Yüksek Lisans

    Türkçe

    Türkçe

    1993

    Elektrik ve Elektronik MühendisliğiHacettepe Üniversitesi

    Elektrik-Elektronik Mühendisliği Ana Bilim Dalı

    DOÇ. DR. TURHAN ÇİFTÇİBAŞI

Geri Dön