Pitch-Normalized Acoustic Features for Robust Children's Speech Recognition
- 17 May 2017
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 24 (8) , 1128-1132
Abstract
In this work, the effectiveness of recently reported SMAC features has been evaluated for robust automatic speech recognition (ASR). The SMAC features consist of normalized first central spectral moments appended with low-order cepstral coefficients. These features have been designed for achieving robustness to both additive noise and the pitch variations. We have explored the SMAC features in severe pitch mismatch ASR task, i.e., decoding of children’s speech on adults’ speech trained ASR system. In those tasks, the SMAC features are still observed to be sensitive to pitch variations. Towards addressing the same, a simple spectral smoothening approach employing adaptive-cepstral truncation is explored prior to the computation of spectral moments. With the proposed modification, the SMAC features are noted to achieve enhanced pitch robustness without affecting their noise immunity. Furthermore, the effectiveness of the proposed features is explored in three dominant acoustic modeling paradigms and varying data conditions. In all the cases, the proposed features are observed to significantly outperform the existing ones.Keywords
This publication has 24 references indexed in Scilit:
- Spectral Moment Features Augmented by Low Order Cepstral Coefficients for Robust ASRIEEE Signal Processing Letters, 2010
- A review of ASR technologies for children's speech2009
- Statistical parametric speech synthesisSpeech Communication, 2009
- Recognition of Noisy Speech Using Dynamic Spectral Subband CentroidsIEEE Signal Processing Letters, 2004
- Robust recognition of children's speechIEEE Transactions on Speech and Audio Processing, 2003
- Creating conversational interfaces for childrenIEEE Transactions on Speech and Audio Processing, 2002
- Voice conversion algorithm based on Gaussian mixture model with dynamic frequency warping of STRAIGHT spectrum2001
- Restructuring speech representations using a pitch-adaptive time–frequency smoothing and an instantaneous-frequency-based F0 extraction: Possible role of a repetitive structure in soundsSpeech Communication, 1999
- Acoustics of children’s speech: Developmental changes of temporal and spectral parametersThe Journal of the Acoustical Society of America, 1999
- Assessment for automatic speech recognition: II. NOISEX-92: A database and an experiment to study the effect of additive noise on speech recognition systemsSpeech Communication, 1993