This work demonstrates the use of the nonlinear time-frequency distribution (NLTFD) of a discrete time energy operator (DTEO) based on amplitude modulation-frequency modulation demodulation techniques as a feature i...This work demonstrates the use of the nonlinear time-frequency distribution (NLTFD) of a discrete time energy operator (DTEO) based on amplitude modulation-frequency modulation demodulation techniques as a feature in speech recognition. The duration distribution based hidden Markov module in a speaker independent large vocabulary mandarin speech recognition system was reconstructed from the feature vectors in the front-end detection stage. The goal was to improve the performance of the existing system by combining new features to the baseline feature vector. This paper also deals with errors associated with using a pre-emphasis filter in the front end processing of the present scheme, which causes an increase in the noise energy at high frequencies above 4 kHz and in some cases degrades the recognition accuracy. The experimental results show that eliminating the pre-emphasis filters from the pre-processing stage and using NLTFD with compensated DTEO combined with Mel frequency cepstrum components give a 21.95% reduction in the relative error rate compared to the conventional technique with 25 candidates used in the test.展开更多
为提高噪声不平稳或不可估的情况下语音识别的稳健性,提出了利用自回归模型和短时平稳性假设,估计干净与噪声环境的语音数据,建立相应的语音识别模型,以达到抗噪效果的稳健语音信号处理方法。在N o iseX-92的4种噪声环境(w h ite,babb l...为提高噪声不平稳或不可估的情况下语音识别的稳健性,提出了利用自回归模型和短时平稳性假设,估计干净与噪声环境的语音数据,建立相应的语音识别模型,以达到抗噪效果的稳健语音信号处理方法。在N o iseX-92的4种噪声环境(w h ite,babb le,vo lvo,destroyer eng ine)从0到20 dB的不同信噪比下的“863”大词汇连续语音标准数据库的平均识别结果表明,该方法能够使得基于段长分布的隐M arkov模型的语音识别系统在25候选时声学层的音节相对错误率下降达到10.85%以下,同时相对正确识别率上升12.13%。展开更多
This work describes an improved feature extractor algorithm to extract the peripheral features of point x(ti,fj) using a nonlinear algorithm to compute the nonlinear time spectrum (NL-TS) pattern. The algo- rithm ob...This work describes an improved feature extractor algorithm to extract the peripheral features of point x(ti,fj) using a nonlinear algorithm to compute the nonlinear time spectrum (NL-TS) pattern. The algo- rithm observes n×n neighborhoods of the point in all directions, and then incorporates the peripheral fea- tures using the Mel frequency cepstrum components (MFCCs)-based feature extractor of the Tsinghua elec- tronic engineering speech processing (THEESP) for Mandarin automatic speech recognition (MASR) sys- tem as replacements of the dynamic features with different feature combinations. In this algorithm, the or- thogonal bases are extracted directly from the speech data using discrite cosime transformation (DCT) with 3×3 blocks on an NL-TS pattern as the peripheral features. The new primal bases are then selected and simplified in the form of the ?dp- operator in the time direction and the ?dp- operator in the frequency di- t f rection. The algorithm has 23.29% improvements of the relative error rate in comparison with the standard MFCC feature-set and the dynamic features in tests using THEESP with the duration distribution-based hid- den Markov model (DDBHMM) based on MASR system.展开更多
In speech recognition systems, the physiological characteristics of the speech production model cause the voiced sections of the speech signal to have an attenuation of approximately 20 dB per decade. Many speech rec...In speech recognition systems, the physiological characteristics of the speech production model cause the voiced sections of the speech signal to have an attenuation of approximately 20 dB per decade. Many speech recognition algorithms have been developed to solve this problem by filtering the input signal with a single-zero high pass filter. Unfortunately, this technique increases the noise energy at high frequencies above 4 kHz, which in some cases degrades the recognition accuracy. This paper solves the problem using a pre-emphasis filter in the front end of the recognizer. The aim is to develop a modified parameterization approach taking into account the whole energy zone in the spectrum to improve the performance of the existing baseline recognition system in the acoustic phase. The results show that a large vocabulary speaker-independent continuous speech recognition system using this approach has a greatly improved recognition rate.展开更多
基金the National High- Tech Research andDevelopm ent Program of China(No. 2 0 0 1AA114 0 71)
文摘This work demonstrates the use of the nonlinear time-frequency distribution (NLTFD) of a discrete time energy operator (DTEO) based on amplitude modulation-frequency modulation demodulation techniques as a feature in speech recognition. The duration distribution based hidden Markov module in a speaker independent large vocabulary mandarin speech recognition system was reconstructed from the feature vectors in the front-end detection stage. The goal was to improve the performance of the existing system by combining new features to the baseline feature vector. This paper also deals with errors associated with using a pre-emphasis filter in the front end processing of the present scheme, which causes an increase in the noise energy at high frequencies above 4 kHz and in some cases degrades the recognition accuracy. The experimental results show that eliminating the pre-emphasis filters from the pre-processing stage and using NLTFD with compensated DTEO combined with Mel frequency cepstrum components give a 21.95% reduction in the relative error rate compared to the conventional technique with 25 candidates used in the test.
文摘为提高噪声不平稳或不可估的情况下语音识别的稳健性,提出了利用自回归模型和短时平稳性假设,估计干净与噪声环境的语音数据,建立相应的语音识别模型,以达到抗噪效果的稳健语音信号处理方法。在N o iseX-92的4种噪声环境(w h ite,babb le,vo lvo,destroyer eng ine)从0到20 dB的不同信噪比下的“863”大词汇连续语音标准数据库的平均识别结果表明,该方法能够使得基于段长分布的隐M arkov模型的语音识别系统在25候选时声学层的音节相对错误率下降达到10.85%以下,同时相对正确识别率上升12.13%。
基金Supported by the National High-Tech Research and Development (863) Program of China (No. 200/AA/14)
文摘This work describes an improved feature extractor algorithm to extract the peripheral features of point x(ti,fj) using a nonlinear algorithm to compute the nonlinear time spectrum (NL-TS) pattern. The algo- rithm observes n×n neighborhoods of the point in all directions, and then incorporates the peripheral fea- tures using the Mel frequency cepstrum components (MFCCs)-based feature extractor of the Tsinghua elec- tronic engineering speech processing (THEESP) for Mandarin automatic speech recognition (MASR) sys- tem as replacements of the dynamic features with different feature combinations. In this algorithm, the or- thogonal bases are extracted directly from the speech data using discrite cosime transformation (DCT) with 3×3 blocks on an NL-TS pattern as the peripheral features. The new primal bases are then selected and simplified in the form of the ?dp- operator in the time direction and the ?dp- operator in the frequency di- t f rection. The algorithm has 23.29% improvements of the relative error rate in comparison with the standard MFCC feature-set and the dynamic features in tests using THEESP with the duration distribution-based hid- den Markov model (DDBHMM) based on MASR system.
基金Supported by the National High- TechnologyDevelopm ent Program of China(No.2 0 0 1AA1140 71)
文摘In speech recognition systems, the physiological characteristics of the speech production model cause the voiced sections of the speech signal to have an attenuation of approximately 20 dB per decade. Many speech recognition algorithms have been developed to solve this problem by filtering the input signal with a single-zero high pass filter. Unfortunately, this technique increases the noise energy at high frequencies above 4 kHz, which in some cases degrades the recognition accuracy. This paper solves the problem using a pre-emphasis filter in the front end of the recognizer. The aim is to develop a modified parameterization approach taking into account the whole energy zone in the spectrum to improve the performance of the existing baseline recognition system in the acoustic phase. The results show that a large vocabulary speaker-independent continuous speech recognition system using this approach has a greatly improved recognition rate.
