Speaker verification in noise using temporal constraints
نویسندگان
چکیده
This paper addresses the problem of state duration modeling in combination with spectral subtraction and Rasta filtering to cancel both additive and convolutional noise in a text-dependent speaker verification task. The results presented in this paper suggest that temporal constraints can lead to reductions of 30 and 14% in the error rates at SNR equal to 0 and 6dB, respectively, without noise canceling techniques. However, with noise canceling methods, temporal restrictions give a lower improvement. The results here shown propose that state duration modeling can be useful in those cases when the noise reduction is low.
منابع مشابه
Speaker dependent temporal constraints combined with speaker independent HMM for speech recognition in noise
This paper addresses the problem of speech recognition in noise using speaker-dependent temporal constraints in the Viterbi algorithm in combination with speaker-independent HMM. It is shown that the speaker-dependent re-estimation of state duration parameters requires a low computational load and a small training database, and can lead to reductions in the error rate as high as 30% or 40% with...
متن کاملUsing timing information in speaker verification
This paper presents an analysis of temporal information as a feature for use in speaker verification systems. The relevance of temporal information in a speaker’s utterances is investigated, both with regard to improving the robustness of modern speaker verification systems and to detecting and deflecting recording attacks. It is shown that the use of timing information provides useful addition...
متن کاملUsing Exciting and Spectral Envelope Information and Matrix Quantization for Improvement of the Speaker Verification Systems
Speaker verification from talking a few words of sentences has many applications. Many methods as DTW, HMM, VQ and MQ can be used for speaker verification. We applied MQ for its precise, reliable and robust performance with computational simplicity. We also used pitch frequency and log gain contour for further improvement of the system performance.
متن کاملUsing Exciting and Spectral Envelope Information and Matrix Quantization for Improvement of the Speaker Verification Systems
Speaker verification from talking a few words of sentences has many applications. Many methods as DTW, HMM, VQ and MQ can be used for speaker verification. We applied MQ for its precise, reliable and robust performance with computational simplicity. We also used pitch frequency and log gain contour for further improvement of the system performance.
متن کاملNoise-robust speaker verification using F0 features
This paper proposes a noise-robust speaker verification method augmented by fundamental frequency (F0). The paper first describes a noise-robust F0 extraction method using the Hough transform. Then, it proposes a robust speaker verification method using multi-stream HMMs which fuse the extracted F0 and cepstral features. Experiments are conducted using fourconnected-digit utterances of Japanese...
متن کامل