Comparison of feature extraction and normalization methods for speaker recognition using grid-audiovisual database

In this paper, a new suggested system for speaker recognition by using hidden markov model (HHM) algorithm. Many researches have been written in this subject, especially by HMM. Arabic language is one of the difficult languages and the work with it is very little, also, the work has been done for text dependent system where HMM is very effective and the algorithm trained at the word level. One the problems in such systems is the noise, so we take it in consideration by adding additive white gaussian noise (AWGN) to the speech signals to see its effect. Here, we used HMM with new algorithm with one state, where two of these components, i.e. (π and A) are removed. This give extremely accelerates the training and testing stages of recognition speeds with lowest memory usage, as seen in the work. The results show an excellent outcome. 100% recognition rate for the tested data, about 91.6% recognition rate with AWGN noise.

Section: Introductionmentioning

confidence: 99%

Arabic speaker recognition using HMM

Hussein

Salman²,

Saeed³

2021

“…In this case, various optical, electronic and radio engineering systems with program control are used. The most famous systems of machine and computer vision with digital image processing [1]- [3]. However, in these papers are set forth general technologies for recognizing objects without solving the problem of sorting minerals by any signs.…”

Section: Introductionmentioning

confidence: 99%

Intelligent machine for sorting semi-precious minerals

Polishchuk

Tkach

Parkhomey

et al. 2021

This paper proposes is devoted to solving the problem of sorting workpieces from minerals and semi-precious stones that are used in jewelry production. Intensive development of production leads to a reduction of natural resources, including semi-precious minerals. Therefore, the task of saving the use of minerals is relevant. The solution is based on the idea a new method for sorting semi-precious minerals is proposed, as well as the design of an intelligent installation for implementing this method. An increase in the accuracy of sorting semiprecious minerals has been achieved through the use of charge-coupled device (CCD) matrices, which automatically determine the volume of each mineral in 3D space. From the simulation work, it was found in comparison with manual sorting, the proposed method has reduced the amount of waste in the manufacture of jewelry by 15% ... 20%. Experimental studies were carried out using the example of such a semiprecious mineral as amber. The cost of raw amber in the amount of one kilogram on the modern market is in the range of 1200…2200 US dollars. The result confirms that the method is superior to the alternative method, with the achieved average waste saving of 20%, the proposed sorting technology makes it possible to obtain a profit of 240...440 US dollars per kilogram of amber. Ultimately, these solutions can improve the efficiency of jewelry production.

“…The mel frequency cepstral coefficients (MFCC) method is a reliable method with high accuracy for high-quality audio recordings [13]. The accuracy rate is more than 90% [14], [15]. The high accuracy of the MFCC method is due to the mel scale which has characteristics similar to human hearing [11].…”

Section: Introductionmentioning

confidence: 99%

“…The characteristics of the mel scale on channels 1 and 2 follow the principle that not all follow a linear pattern, so that each channel follows a linear and exponential pattern. The frequency is then on the mel scale with (15). The width of the mel scale plane of channel 1 and channel 2 is shown in Figure 5.…”

mentioning

confidence: 99%

Feature extraction with mel scale separation method on noise audio recordings

Huizen

Kurniati

2021

This paper focuses on improving the accuracy of noise audio recordings. High-quality audio recording, extraction using the mel frequency cepstral coefficients (MFCC) method produces high accuracy. While the low-quality is because of noise, the accuracy is low. Improved accuracy by investigating the effect of bandwidth on the mel scale. The proposed improvement uses the mel scale separation methods into two frequency channels (MFCC dual-channel). For the comparison method using the mel scale bandwidth without separation (MFCC single-channel). Feature analysis using k-mean clustering. The data uses a noise variance of up to -16 dB. Testing on the MFCC single-channel method for -16 dB noise has an accuracy of 47.5%, while the MFCC dual-channel method has an accuracy better of 76.25%. The next test used adaptive noise-canceling (ANC) to reduce noise before extraction. The result is that the MFCC single-channel method has an accuracy of 82.5% and the MFCC dual-channel method has an accuracy better of 83.75%. High-quality audio recording testing for the MFCC single-channel method has an accuracy of 92.5% and the MFCC dual-channel method has an accuracy better of 97.5%. The test results show the effect of mel scale bandwidth to increase accuracy. The MFCC dual-channel method has higher accuracy.