کد مقاله کد نشریه سال انتشار مقاله انگلیسی نسخه تمام متن
563117 875471 2013 28 صفحه PDF دانلود رایگان
عنوان انگلیسی مقاله ISI
Importance sampling to compute likelihoods of noise-corrupted speech
موضوعات مرتبط
مهندسی و علوم پایه مهندسی کامپیوتر پردازش سیگنال
پیش نمایش صفحه اول مقاله
Importance sampling to compute likelihoods of noise-corrupted speech
چکیده انگلیسی

One way of making speech recognisers more robust to noise is model compensation. Rather than enhancing the incoming observations, model compensation techniques modify a recogniser's state-conditional distributions so they model the speech in the target environment. Because the interaction between speech and noise is non-linear, even for Gaussian speech and noise the corrupted speech distribution has no closed form. Thus, model compensation methods approximate it with a parametric distribution, such as a Gaussian or a mixture of Gaussians. The impact of this approximation has never been quantified. This paper therefore introduces a non-parametric method to compute the likelihood of a corrupted speech observation. It uses sampling and, given speech and noise distributions and a mismatch function, is exact in the limit. It therefore gives a theoretical bound for model compensation. Though computing the likelihood is computationally expensive, the novel method enables a performance comparison based on the criterion that model compensation methods aim to minimise: the KL divergence to the ideal compensation. It gives the point where the Kullback–Leibler (KL) divergence is zero. This paper examines the performance of various compensation methods, such as vector Taylor series (VTS) and data-driven parallel model combination (DPMC). It shows that more accurate modelling than Gaussian-for-Gaussian compensation improves the performance of speech recognition.


► Model compensation methods approximate the noise-corrupted speech.
► They normally produce parametric distributions, like Gaussians, but the corrupted speech has no closed form.
► A new sampling method in the limit computes the exact corrupted-speech likelihood.
► It is then possible to quantify how close to optimal model compensation methods are.

ناشر
Database: Elsevier - ScienceDirect (ساینس دایرکت)
Journal: Computer Speech & Language - Volume 27, Issue 1, January 2013, Pages 322–349
نویسندگان
, ,