A unique approach in text independent speaker recognition using MFCC feature sets and probabilistic neural network

Ahmad, Khan Suhail; Thosar, Anil S.; Nirmal, Jagannath; Pande, Vinay

doi:10.1109/icapr.2015.7050669

Cited by 60 publications

(25 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The researcher in work [8] used the same concepts which are used in work [6], and they are focus on the problem of finding a trade-off between the time duration of speech sample and the probability of error. [9] presents an FSR scheme which use a combination of MFCC and its delta derivatives DMFCC and DDMFCC. also, the probabilistic neural network (PNN) in the modeling domain is used to achieve lower operational times during the training steps.…”

Section: Related Workmentioning

confidence: 99%

Arabic Speech Recognition Based on Knn, J48, and LVQ

Alwahed

Jawad

2019

Iraqi Journal of ICT

View full text Add to dashboard Cite

Most systems of speaker recognition work on speech feature primarily classified of being a low level which considerably relies on speaker physical characteristics and, to the lower extent, the acquired speaking habits. In this paper present a system to recognition and identification in Arabic speaker. It includes two phases (training phase and testing phase) each phase includes the using of audio features (Mean, Standard Division, Zero Crossing, Amplitude). after get the feature, the recognition step is using (J48, KNN, LVQ),) where the Nearest Neighbor (KNN) applied o get the similarity of the data training and data testing , LVQ neural network used for Speech Recognition and Arabic language Identification. This sentence contains words especially kidnappings and kidnappers are ten sentences and pronounce these sentences by 10 people, five men and five women of different ages and each of the ten pronunciation of all sentences, so a total of 100 samples and the samples were recorded on audio and wave. The results of the sentences pronounced by women are higher than the results of the same sentences pronounced by men. They achieved better recognition rate 85, 93, 96.4%

show abstract

Section: Related Workmentioning

confidence: 99%

Arabic Speech Recognition Based on Knn, J48, and LVQ

Alwahed

Jawad

2019

Iraqi Journal of ICT

View full text Add to dashboard Cite

show abstract

“…The features should be robust against noise, should have low variability for sessions of same speaker and large between-speaker variability [2], [7]. MFCCs are effective features for speaker identification.…”

Section: Related Workmentioning

confidence: 99%

“…In [12], fractional Fourier transform based modified MFCC features are proposed for speaker recognition. A comparison between MFCC, Delta-MFCC and Delta-Delta MFCC for number of filter coefficients is carried out in [7]. A similar comparison for MFCC and Delta-MFCC is carried out for mean and variance normalization in [13].…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Combining dynamic features with MFCC for text-independent speaker identification

Chaudhari

Rahulkar

Dhonde

2015

2015 International Conference on Information Processing (ICIP)

View full text Add to dashboard Cite

This paper presents text-independent speaker identification scheme based on the combination of dynamic features with Mel Frequency Cepstral Coefficients (MFCC). First, MFCCs are used to extract the speaker specific features and then vector quantization is used to model the speaker. We compare the performance of speaker identification system using number of MFCC filters and centroids of vector quantization. In addition, we also compare the effect of possible combinations of MFCC with log energy, delta and double delta coefficients for number of filters and centroids of vector quantization on recognition rate. The experimental results have been evaluated on the developed database of 75 speakers.

show abstract

“…Wavelet analysis and the effect of number of MFCC features on recognition accuracy has been proposed in [8]. The effect of feature vector size of MFCC on identification accuracy is presented in [9]. The number of choices in creating feature vectors from MFCC has been assessed in [10].In this paper, the effect of number of MFCC filters on recognition rate is studied.…”

Section: Introductionmentioning

confidence: 99%

Effect of Varying MFCC Filters for Speaker Recognition

Chaudhari¹,

Dhonde²

2015

IJCA

View full text Add to dashboard Cite

This paper presents speaker recognition system with emphasis on MFCC feature extraction scheme. The optimum number of MFCC filter selection is necessary for the performance of speaker recognition. In this paper, the number of MFCC filters are varied. The effect of varying filters on computational time required for training and testing phase is provided in this paper. The experimental results have been evaluated on the developed database of 75 speakers. The recognition rate achieved is 96% in case of 30 MFCC filters with approximate computational time (testing phase) of 79.66 seconds.

show abstract

A unique approach in text independent speaker recognition using MFCC feature sets and probabilistic neural network

Cited by 60 publications

References 20 publications

Arabic Speech Recognition Based on Knn, J48, and LVQ

Arabic Speech Recognition Based on Knn, J48, and LVQ

Combining dynamic features with MFCC for text-independent speaker identification

Effect of Varying MFCC Filters for Speaker Recognition

Contact Info

Product

Resources

About