A Mixture Model Approach for Formant Tracking and the Robustness of Student's-t Distribution

• Published in: IEEE transactions on audio, speech, and language processing speech, and language processing, 2012-12, Vol.20 (10), p.2626-2636
• Main Authors: Sundar, Harshavardhan, Seelamantula, Chandra Sekhar, Sreenivas, Thippur V.
• Format: Article
• Language: English
• Subjects: Adaptive filters; Additives; Amplitude modulation; Applied sciences; Computational modeling; Computer simulation; Contours; Detection, estimation, filtering, equalization, prediction; Estimating techniques; Exact sciences and technology; Formant tracking; Frequency modulation; Gaussian mixture model; Gaussian mixture model (GMM); Hidden Markov models; Information, signal and communications theory; Materials handling; Modulation, demodulation; multimodal density estimation; Noise; Robustness; Signal and communications theory; Signal, noise; Speech; statistical mixture modeling; Student's-t mixture model (tMM); Studies; Telecommunications and information theory; Tracking
• ISSN: 1558-7916; 2329-9290; 1558-7924; 2329-9304
• DOI: 10.1109/TASL.2012.2209418

We address the problem of robust formant tracking in continuous speech in the presence of additive noise. We propose a new approach based on mixture modeling of the formant contours. Our approach consists of two main steps: (i) Computation of a pyknogram based on multiband amplitude-modulation/frequency-modulation (AM/FM) decomposition of the input speech; and (ii) Statistical modeling of the pyknogram using mixture models. We experiment with both Gaussian mixture model (GMM) and Student's-t mixture model (tMM) and show that the latter is robust with respect to handling outliers in the pyknogram data, parameter selection, accuracy, and smoothness of the estimated formant contours. Experimental results on simulated data as well as noisy speech data show that the proposed tMM-based approach is also robust to additive noise. We present performance comparisons with a recently developed adaptive filterbank technique proposed in the literature and the classical Burg's spectral estimator technique, which show that the proposed technique is more robust to noise.