An experimental comparison of modelling techniques for speaker recognition under limited data condition

Jayanna, H. S.; Prasanna, S. R. Mahadeva

doi:10.1007/s12046-009-0042-9

Cited by 13 publications

(11 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It can be implied from that study that all popular speaker recognition techniques, normally trained using substantial speech data, will suffer from similar performance degradation when trained and tested by using limited speech data. Thus, the proposed system presented in this paper shows better limited data condition performance than all the traditional methods described by Jayanna and Prasanna (2009).…”

Section: Resultsmentioning

confidence: 80%

Speaker verification using heterogeneous neural network architecture with linear correlation speech activity detection

2013

View full text Add to dashboard Cite

This paper presents a multi-level speaker verification system that uses 64 discrete Fourier transform spectrum components as input feature vectors. A speech activity detection technique is used as a pre-processing stage to identify vowel phoneme boundaries within a speech sample. A modified self-organising map (SOM) is then used to filter the speech data by using cluster information extracted from three vowels for a claimed speaker. This SOM filtering stage also provides coarse speaker verification. Finally, a second speaker verification level of three multi-layer perceptron networks classifies the filtered frames provided by the SOMs. These multi-layer perceptrons work as fine-grained vowel-based speaker verifiers. The proposed verification algorithm shows a performance of 94.54% when evaluated using 50 speakers from the Centre for Spoken Language Understanding speaker verification database. In addition, it is shown that the novel discrete Fourier transform spectrum-based linear correlation pre-processing technique, presented here, provides the system with greater robustness against changes in speech volume levels when compared with an equivalent energy frame analysis.

show abstract

Section: Resultsmentioning

confidence: 80%

Speaker verification using heterogeneous neural network architecture with linear correlation speech activity detection

2013

View full text Add to dashboard Cite

show abstract

“…A limited amount of speech data (20-30 msec for training and 8-15 msec for testing) is used for evaluation. Kmeans clustering algorithm based on iterative refinement approach is used for formation of speaker model, which is effective specifically for limited data condition [20]. The system performance is measured in terms of Percentage Identification Accuracy (PIA) given as, the ratio of number of correctly classified test sequences to the total number of test sequences calculated in percentage.…”

Section: Resultsmentioning

confidence: 99%

Speaker Recognition in Mismatch Conditions: A Feature Level Approach

Chougule¹,

Chavan²

2017

IJIGSP

View full text Add to dashboard Cite

Abstract-Mismatch in speech data is one of the major reasons limiting the use of speaker recognition technology in real world applications. Extracting speaker specific features is a crucial issue in the presence of noise and distortions. Performance of speaker recognition system depends on the characteristics of extracted features. Devices used to acquire the speech as well as the surrounding conditions in which speech is collected, affects the extracted features and hence degrades the decision rates. In view of this, a feature level approach is used to analyze the effect of sensor and environment mismatch on speaker recognition performance. The goal here is to investigate the robustness of segmental features in speech data mismatch and degradation. A set of features derived from filter bank energies namely: Mel Frequency Cepstral Coefficients (MFCCs), Linear Frequency Cepstral Coefficients (LFCCs), Log Filter Bank Energies (LOGFBs) and Spectral Subband Centroids (SSCs) are used for evaluating the robustness in mismatch conditions. A novel feature extraction technique named as Normalized Dynamic Spectral Features (NDSF) is proposed to compensate the sensor and environment mismatch. A significant enhancement in recognition results is obtained with proposed feature extraction method.

show abstract

“…As we know, the number of feature vectors should be not less than ten times the number of nonoverlapping clusters (Jayanna and Mahadeva, 2009). …”

Section: Discussionmentioning

confidence: 99%

Random vector quantisation modelling in automatic speaker verification

Djellali

Laskri

2013

IJBM

View full text Add to dashboard Cite

Automatic speaker verification (ASV) is a binary classification task. It consists of accepting or rejecting the claimed identity. ASV system has to decide whether a claimed speaker uttered a sentence. This paper proposes an algorithm called impostor random vector quantisation (IRVQ) based on multiples random codebook. IRVQ represents the impostor model also called universal background model (UBM) and we compare it to the second algorithm called partial impostor VQ (IVQ) for vector quantisation (VQ) in speaker verification. The present study demonstrates that the several random selected codebooks representing impostor models give better results and less half total error HTER than impostor IVQ method and baseline system. The performance of these models is evaluated on Arabic speaker verification dataset. However, this improvement also depends on the codebook size. The assumption concerning partitioning impostor acoustic spaces and choosing the best subspace to represent impostor model specific to each speaker is an efficient approach.

show abstract

An experimental comparison of modelling techniques for speaker recognition under limited data condition

Cited by 13 publications

References 15 publications

Speaker verification using heterogeneous neural network architecture with linear correlation speech activity detection

Speaker verification using heterogeneous neural network architecture with linear correlation speech activity detection

Speaker Recognition in Mismatch Conditions: A Feature Level Approach

Random vector quantisation modelling in automatic speaker verification

Contact Info

Product

Resources

About