Accuracy of speaker verification via orthogonal parameters for noisy speech

Shridhar, M.; Baraniecki, M.

doi:10.1109/icassp.1979.1170778

Cited by 3 publications

(3 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Time-delay neural networks have been used successfully in both speech recognition and speaker recognition. Linear predictive parameters and their derived parameters related to speaker's vocal tract have been used for speaker identification system [2,4,5]. The linear predictive coding (LPC) derived parameter with hidden Markov models has been used in both speech and speaker identification system [6,7].…”

Section: Related Workmentioning

confidence: 99%

Utterance Based Speaker Identification Using ANN

Das¹

2014

IJCSEA

View full text Add to dashboard Cite

show abstract

Section: Related Workmentioning

confidence: 99%

Utterance Based Speaker Identification Using ANN

Das¹

2014

IJCSEA

View full text Add to dashboard Cite

show abstract

“…The verification algorithm outlined in the pre- However, the above procedure was found to be very accurate when reference speech utterances were also obtained under the same conditions as test utterance [6].…”

Section: Verification Without Preprocessingmentioning

confidence: 99%

“…For analysis, each utterance was divided into 100 frames to provide linear time normalization, a procedure found necessary in the authors' previous study [6], silent portions were eliminated only at the beginning and end.…”

Section: ) Papa Needs Two Singersmentioning

confidence: 99%

A speaker verification algorithm for speech utterances corrupted by noise with unknown statistics

Baraniecki¹,

Shridhar²

ICASSP '80. IEEE International Conference on Acoustics, Speech, and Signal Processing

Self Cite

View full text Add to dashboard Cite

of these algorithms.The goal of this study was to develop a speaker verification algorithm which yielded high accuracy verification when test utterances were obtained under different noise conditions. Tn this study, test utterances were obtained i) fron a noisy telephone channel ii) with additive white noise iii) fron recordings in rooms with typical ambient noise.Initially the verification algorithm utilizing orthogonal parameters derived fron time normalized speech without preprocessing were used to evaluate the accuracies obtainable. Our studies indicated that accuracies were limited to about 45%.The test utterances obtained above were then processed by the adaptikre noise cancelling algorithm to filter out the distortion. The verification algorithm, when applied to the processed speech yielded a significant improvement in verification accuracy to the extent of about 95%.

show abstract