Noise-robust speech triage

Noise-robust speech triage

WoS Article

A method is presented in which conventional speech algorithms are applied, with no modifications,to improve their performance in extremely noisy environments. It has been demonstrated that, foreigen-channel algorithms, pre-training multiple speaker identification (SID) models at a lattice ofsignal-to-noise-ratio (SNR) levels and then performing SID using the appropriate SNR dependentmodel was successful in mitigating noise at all SNR levels. In those tests, it was found that SID performancewas optimized when the SNR of the testing and training data were close or identical. Inthis current effort multiple i-vector algorithms were used, greatly improving both processingthroughput and equal error rate classification accuracy. Using identical approaches in the samenoisy environment, performance of SID, language identification, gender identification, and diarizationwere significantly improved. A critical factor in this improvement is speech activity detection(SAD) that performs reliably in extremely noisy environments, where the speech itself is barelyaudible. To optimize SAD operation at all SNR levels, two algorithms were employed. The firstmaximized detection probability at low levels (10 dB  SNR < 10 dB) using just the voicedspeech envelope, and the second exploited features extracted from the original speech to improveoverall accuracy at higher quality levels (SNR10 dB).

A method is presented in which conventional speech algorithms are applied, with no modifications,to improve their performance in extremely noisy environments. It has been demonstrated that, foreigen-channel algorithms, pre-training multiple speaker identification (SID) models at a lattice ofsignal-to-noise-ratio (SNR) levels and then performing SID using the appropriate SNR dependentmodel was successful in mitigating noise at all SNR levels. In those tests, it was found that SID performancewas optimized when the SNR of the testing and training data were close or identical. Inthis current effort multiple i-vector algorithms were used, greatly improving both processingthroughput and equal error rate classification accuracy. Using identical approaches in the samenoisy environment, performance of SID, language identification, gender identification, and diarizationwere significantly improved. A critical factor in this improvement is speech activity detection(SAD) that performs reliably in extremely noisy environments, where the speech itself is barelyaudible. To optimize SAD operation at all SNR levels, two algorithms were employed. The firstmaximized detection probability at low levels (10 dB  SNR < 10 dB) using just the voicedspeech envelope, and the second exploited features extracted from the original speech to improveoverall accuracy at higher quality levels (SNR10 dB).

speech algorithms, noisy environments, multiple speaker identification

speech algorithms, noisy environments, multiple speaker identification

BARTOS, A.; CIPR, T.; NELSON, D.; SCHWARZ, P.; BANOWETZ, J.; JERABEK, L.

2019

23.04.2018

1520-8524

JOURNAL OF THE ACOUSTICAL SOCIETY OF AMERICA

143

4

United States of America

2313

2320

8

https://asa.scitation.org/doi/10.1121/1.5031029

A_L_Bartos_JASMAN_vol_143iss_42313_1