Publications

Refine Results

(Filters Applied) Clear All

Tagged As

Magnitude-only estimation of handset nonlinearity with application to speaker recognition

May 15, 1998
  |
Conference Paper
Author:
Thomas F. Quatieri
…
Published in:
Proc. of the 1998 IEEE Int. Conf. on Acoustics, Speech and Signal Processing, ICASSP, Vol. II, Speech Processing II; Neural Networks for Signal Processing, 12-15 May 1998, pp. 745-748.
Topic:
speaker recognition
R&D area:
R&D group:

Summary

A method is described for estimating telephone handset nonlinearity by matching the spectral magnitude of the distorted signal to the output of a nonlinear channel model, driven by an undistorted reference. The "magnitude-only" representation allows the model to directly match unwanted speech formants that arise over nonlinear channels and that are a potential source of degradation in speaker and speech recognition algorithms. As such, the method is particularly suited to algorithms that use only spectral magnitude information. The distortion model consists of a memoryless polynomial nonlinearity sandwiched between two finite-length linear filters. Minimization of a mean-squared spectral magnitude error, with respect to model parameters, relies on iterative estimation via a gradient descent technique, using a Jacobian in the iterative correction term with gradients calculated by finite-element approximation. Initial work has demonstrated the algorithm's usefulness in speaker recognition over telephone channels by reducing mismatch between high- and low-quality handset conditions.
READ LESS

Summary

A method is described for estimating telephone handset nonlinearity by matching the spectral magnitude of the distorted signal to the output of a nonlinear channel model, driven by an undistorted reference. The "magnitude-only" representation allows the model to directly match unwanted speech formants that arise over nonlinear channels and that...

READ MORE
Magnitude-only estimation of handset nonlinearity with application to speaker recognition

Comparison of background normalization methods for text-independent speaker verification

September 22, 1997
  |
Conference Paper
Author:
Douglas A. Reynolds
Published in:
5th European Conf. on Speech Communication and Technology, EUROSPEECH, 22-25 September 1997.
Topic:
speaker recognition
R&D area:
R&D group:

Summary

This paper compares two approaches to background model representation for a text-independent speaker verification task using Gaussian mixture models. We compare speaker-dependent background speaker sets to the use of a universal, speaker-independent background model (UBM). For the UBM, we describe how Bayesian adaptation can be used to derive claimant speaker models, providing a structure leading to significant computational savings during recognition. Experiments are conducted on the 1996 NIST Speaker Recognition Evaluation corpus and it is clearly shown that a system using a UBM and Bayesian adaptation of claimant models produces superior performance compared to speaker-dependent background sets or the UBM with independent claimant models. In addition, the creation and use of a telephone handset-type detector and a procedure called hnorm is also described which shows further, large improvements in verification performance, especially under the difficult mismatched handset conditions. This is believed to be the first use of applying a handset-type detector and explicit handset-type normalization for the speaker verification task.
READ LESS

Summary

This paper compares two approaches to background model representation for a text-independent speaker verification task using Gaussian mixture models. We compare speaker-dependent background speaker sets to the use of a universal, speaker-independent background model (UBM). For the UBM, we describe how Bayesian adaptation can be used to derive claimant speaker...

READ MORE
Comparison of background normalization methods for text-independent speaker verification

Improving wordspotting performance with artificially generated data

May 9, 1996
  |
Conference Paper
Author:
Eric I. Chang
…
Published in:
Proc. IEEE Int. Conf. on Acoustics, Speech, and Signal Processing, ICASSP, 9 May 1996, pp. 526-9.
Topic:
speaker recognition
R&D area:
R&D group:

Summary

Lack of training data is a major problem that limits the performance of speech recognizers. Performance can often only be improved by expensive collection of data from many different talkers. This paper demonstrates that artificially transformed speech can increase the variability of training data and increase the performance of a wordspotter without additional expensive data collection. This approach was shown to be effective on a high-performance whole-word wordspotter on the Switchboard Credit Card database. The proposed approach used in combination with a discriminative training approach increased the Figure of Merit of the wordspotting system by 9.4% percentage points (62.5% to 71.9%). The increase in performance provided by artificially transforming speech was roughly equivalent to the increase that would have been provided by doubling the amount of training data. The performance of the wordspotter was also compared to that of human listeners who were able to achieve lower error rates because of improved consonant recognition.
READ LESS

Summary

Lack of training data is a major problem that limits the performance of speech recognizers. Performance can often only be improved by expensive collection of data from many different talkers. This paper demonstrates that artificially transformed speech can increase the variability of training data and increase the performance of a...

READ MORE
Improving wordspotting performance with artificially generated data

Fine structure features for speaker identification

May 7, 1996
  |
Conference Paper
Author:
Charles R. Jankowski Jr
…
Published in:
Proc. IEEE Int. Conf. on Acoustics, Speech and Signal Processing, ICASSP, Vol. 2, Speech (Part II), 7-10 May 1996, pp. 689-692.
Topic:
speaker recognition
R&D area:
R&D group:

Summary

The performance of speaker identification (SID) systems can be improved by the addition of the rapidly varying "fine structure" features of formant amplitude and/or frequency modulation and multiple excitation pulses. This paper shows how the estimation of such fine structure features can be improved further by obtaining better estimates of formant frequency locations and uncovering various sources of error in the feature extraction systems. Most female telephone speech showed "spurious" formants, due to distortion in the telephone network. Nevertheless, SID performance was greatest with these spurious formants as formant estimates. A new feature has also been identified which can increase SID performance: cepstral coefficients from noise in the estimated excitation waveform. Finally, statistical tools have been developed to explore the relative importance of features used for SID, with the ultimate goal of uncovering the source of the features that provide SID performance improvement.
READ LESS

Summary

The performance of speaker identification (SID) systems can be improved by the addition of the rapidly varying "fine structure" features of formant amplitude and/or frequency modulation and multiple excitation pulses. This paper shows how the estimation of such fine structure features can be improved further by obtaining better estimates of...

READ MORE
Fine structure features for speaker identification

The effects of handset variability on speaker recognition performance: experiments on the switchboard corpus

May 7, 1996
  |
Conference Paper
Author:
Douglas A. Reynolds
Published in:
Proc. IEEE Int. Conf. on Acoustics, Speech and Signal Processing, ICASSP, Vol. 1, 7-10 May 1996, pp. 113-116.
Topic:
speaker recognition
R&D area:
R&D group:

Summary

This paper presents an empirical study of the effects of handset variability on text-independent speaker recognition performance using the Switchboard corpus. Handset variability occurs when training speech is collected using one type of handset, but a different handset is used for collecting test speech. For the Switchboard corpus, the calling telephone number associated with a file is used to imply the handset used. Analysis of experiments designed to focus on handset variability on the SPIDRE database and the May95 NIST speaker recognition evaluation database, show that a performance gap between matched and mismatched handset tests persists even after applying several standard channel compensation techniques. Error rates for the mismatched tests are over 4 times those for the matched tests. Lastly, a new energy dependent cepstral mean subtraction technique is proposed to compensate for nonlinear distortions, but is not found to improve performance on the databases used.
READ LESS

Summary

This paper presents an empirical study of the effects of handset variability on text-independent speaker recognition performance using the Switchboard corpus. Handset variability occurs when training speech is collected using one type of handset, but a different handset is used for collecting test speech. For the Switchboard corpus, the calling...

READ MORE
The effects of handset variability on speaker recognition performance: experiments on the switchboard corpus

Unsupervised topic clustering of switchboard speech messages

May 7, 1996
  |
Conference Paper
Author:
Beth Carlson Kittel
Published in:
Proc. IEEE Int. Conf. on Acoustics, Speech and Signal Processing, ICASSP, Vol. 1, 7-10 May 1996, pp. 315-318.
Topic:
speaker recognition
R&D area:
R&D group:

Summary

This paper presents a statistical technique which can be used to automatically group speech data records based on the similarity of their content. A tree-based clustering algorithm is used to generate a hierarchical structure for the corpus. This structure can then be used to guide the search for similar material in data from other corpora. The SWITCHBOARD Speech Corpus was used to demonstrate these techniques, since it provides sets of speech files which are nominally on the same topic. Excellent automatic clustering was achieved on the truth text transcripts provided with the SWITCHBOARD corpus, with an average cluster purity of 97.3%. Degraded clustering was achieved using the output transcriptions of a speech recognizer, with a clustering purity of 61.4%.
READ LESS

Summary

This paper presents a statistical technique which can be used to automatically group speech data records based on the similarity of their content. A tree-based clustering algorithm is used to generate a hierarchical structure for the corpus. This structure can then be used to guide the search for similar material...

READ MORE
Unsupervised topic clustering of switchboard speech messages

Measuring fine structure in speech: application to speaker identification

May 9, 1995
  |
Conference Paper
Author:
Charles R. Jankowski Jr
…
Published in:
Proc. IEEE Int. Conf. on Acoustics, Speech and Signal Processing, ICASSP, Vol. 1, 9-12 May 1995, pp. 325-328.
Topic:
speaker recognition
R&D area:
R&D group:

Summary

The performance of systems for speaker identification (SID) can be quite good with clean speech, though much lower with degraded speech. Thus it is useful to search for new features for SID, particularly features that are robust over a degraded channel. This paper investigates features that are based on amplitude and frequency modulations of speech formants, high resolution measurement of fundamental frequency and location of "secondary pulses," measured using a high-resolution energy operator. When these features are added to traditional features using an existing SID system with a 168 speaker telephone speech database, SID performance improved by as much as 4% for male speakers and 8.2% for female speakers.
READ LESS

Summary

The performance of systems for speaker identification (SID) can be quite good with clean speech, though much lower with degraded speech. Thus it is useful to search for new features for SID, particularly features that are robust over a degraded channel. This paper investigates features that are based on amplitude...

READ MORE
Measuring fine structure in speech: application to speaker identification

The effects of telephone transmission degradations on speaker recognition performance

May 9, 1995
  |
Conference Paper
Author:
Douglas A. Reynolds
…
Published in:
Proc. IEEE Int. Conf. on Acoustics, Speech and Signal Processing, ICASSP, Vol. 1, Speech, 9-12 May 1995, pp. 329-332.
Topic:
speaker recognition
R&D area:
R&D group:

Summary

The two largest factors affecting automatic speaker identification performance are the size of the population an the degradations introduced by noisy communication, channels (e.g., telephone transmission). To examine experimentally these two factors, this paper presents text-independent speaker identification results for varying speaker population sizes up to 630 speakers for both clean, wideband speech and telephone speech. A system based on Gaussian mixture speaker identification and experiments are conducted on the TIMIT and NTIMIT databases. This is believed to be the first speaker identification experiments on the complete 630 speaker TIMIT and NTIMIT databases and the largest text-independent speaker identification task reported to date. Identification accuracies of 99.5% and 60.7% are achieved on the TIMIT and NTIMIT databases, respectively. This paper also presents experiments which examine and attempt to quantify the performance loss associated with various telephone degradations by systematically degrading the TIMIT speech in a manner consistent with measured NTIMIT degradations and measuring the performance loss at each step. It is found that the standard degradations of filtering and additive noise do not account for all of the performance gap between the TIMIT and NTIMIT data. Measurements of nonlinear microphone distortions are also...
READ LESS

Summary

The two largest factors affecting automatic speaker identification performance are the size of the population an the degradations introduced by noisy communication, channels (e.g., telephone transmission). To examine experimentally these two factors, this paper presents text-independent speaker identification results for varying speaker population sizes up to 630 speakers for both...

READ MORE
The effects of telephone transmission degradations on speaker recognition performance

Large population speaker identification using clean and telephone speech

March 1, 1995
  |
Journal Article
Author:
Douglas A. Reynolds
Published in:
IEEE Signal Process. Lett., Vol. 2, No. 3, March 1995, pp. 46-48.
Topic:
speaker recognition
R&D area:
R&D group:

Summary

This paper presents text-independent speaker identification results for varying speaker population sizes up to 630 speakers for both clean, wideband speech, and telephone speech. A system based on Gaussian mixture speaker models is used for speaker identification, and experiments are conducted on the TIMIT and NTIMIT databases. The TIMIT results show large population performance under near-ideal conditions, and the NTIMIT results show the corresponding accuracy loss due to telephone transmission. These are believed to be the first speaker identification experiments on the complete 630 speaker TIMIT and NTIMIT databases and the largest text-independent speaker identification task reported to date. Identification accuracies of 99.5 and 60.7% were achieved on the TIMIT and NTIMIT databases, respectively.
READ LESS

Summary

This paper presents text-independent speaker identification results for varying speaker population sizes up to 630 speakers for both clean, wideband speech, and telephone speech. A system based on Gaussian mixture speaker models is used for speaker identification, and experiments are conducted on the TIMIT and NTIMIT databases. The TIMIT results...

READ MORE
Large population speaker identification using clean and telephone speech

Robust text-independent speaker identification using Gaussian mixture speaker models

January 1, 1995
  |
Journal Article
Author:
Douglas A. Reynolds
…
Published in:
IEEE Trans. Speech Audio Process., Vol. 3, No. 1, January 1995, pp. 72-83.
Topic:
speaker recognition
R&D area:
R&D group:

Summary

This paper introduces and motivates the use of Gaussian mixture models (GMM) for robust text-independent speaker identification. The individual Gaussian components of a GMM are shown to represent some general speaker-dependent spectral shapes that are effective for modeling speaker identify. The focus of this work is on applications which require high identification rates using short utterance from unconstrained conversational speech and robustness to degradations produced by transmission over a telephone channel. A complete experimental evaluation of the Gaussian mixture speaker model is conducted on a 49 speaker, conversational telephone speech database. The experiments examine algorithmic issues (initializations, variance limiting, model order selection), spectral variability robustness techniques, large population performance, and comparisons to other speaker modeling techniques (uni-modal Gaussian, VQ codebook, tied Gaussian mixture, and radial basis functions). The Gaussian mixture speaker model attains 96.8% identification accuracy using 5 second clean speech utterances and 80.8% accuracy using 15 second telephone speech utterances with a 49 speaker population and is shown to outperform the other speaker modeling techniques on an identical 16 speaker telephone speech task.
READ LESS

Summary

This paper introduces and motivates the use of Gaussian mixture models (GMM) for robust text-independent speaker identification. The individual Gaussian components of a GMM are shown to represent some general speaker-dependent spectral shapes that are effective for modeling speaker identify. The focus of this work is on applications which require...

READ MORE
Robust text-independent speaker identification using Gaussian mixture speaker models
121-130 of 140