Structural maximum a posteriori linear regression for fast HMM adaptation

SUMMARYStatistical speech recognition using continuous-density hidden Markov models (CDHMMs) has yielded many practical applications. However, in general, mismatches between the training data and input data significantly degrade recognition accuracy. Various acoustic model adaptation techniques using a few input utterances have been employed to overcome this problem. In this article, we survey these adaptation techniques, including maximum a posteriori (MAP) estimation, maximum likelihood linear regression (MLLR), and eigenvoice. We also present a schematic view called the adaptation pyramid to illustrate how these methods relate to each other. key words: speech recognition, acoustic model adaptation, hidden Markov models

Section: Structural Approachmentioning

confidence: 99%

Acoustic Model Adaptation for Speech Recognition

Sairyo

2010

“…These transformation parameters are estimated based on structural MAP criterion [19] in which the structure of the decision tree is utilized for robust estimation. In the estimation process, the transformation parameters of the parent node are used as the parameters of the prior distribution of the transformation parameters to be estimated [20].…”

Section: Simultaneous Adaptation Of Speaker and Style From Average Vomentioning

confidence: 99%

HMM-Based Style Control for Expressive Speech Synthesis with Arbitrary Speaker's Voice Using Model Adaptation

Nose

Tachibana

Kobayashi

2009

Takashi NOSE†a) , Student Member, Makoto TACHIBANA †b) , and Takao KOBAYASHI †c) , Members SUMMARY This paper presents methods for controlling the intensity of emotional expressions and speaking styles of an arbitrary speaker's synthetic speech by using a small amount of his/her speech data in HMMbased speech synthesis. Model adaptation approaches are introduced into the style control technique based on the multiple-regression hidden semiMarkov model (MRHSMM). Two different approaches are proposed for training a target speaker's MRHSMMs. The first one is MRHSMM-based model adaptation in which the pretrained MRHSMM is adapted to the target speaker's model. For this purpose, we formulate the MLLR adaptation algorithm for the MRHSMM. The second method utilizes simultaneous adaptation of speaker and style from an average voice model to obtain the target speaker's style-dependent HSMMs which are used for the initialization of the MRHSMM. From the result of subjective evaluation using adaptation data of 50 sentences of each style, we show that the proposed methods outperform the conventional speaker-dependent model training when using the same size of speech data of the target speaker. key words: expressive speech, HMM-based speech synthesis, style control, multiple-regression HSMM (MRHSMM), model adaptation, average voice model

“…To address this problem, maximum a posteriori (MAP) [4] based linear regression, which estimates the transformation matrixes using the MAP criterion, has been proposed. Maximum a posteriori linear regression (MAPLR) [5]- [9] and structural MAPLR (SMAPLR) [10] are notable examples. These approaches incorporate the prior knowledge to address the potential over-fitting problem and have been shown to be successful.…”

Section: Introductionmentioning

confidence: 99%

Smoothing Method for Improved Minimum Phone Error Linear Regression

Pan

et al. 2014

SUMMARYA smoothing method for minimum phone error linear regression (MPELR) is proposed in this paper. We show that the objective function for minimum phone error (MPE) can be combined with a prior mean distribution. When the prior mean distribution is based on maximum likelihood (ML) estimates, the proposed method is the same as the previous smoothing technique for MPELR. Instead of ML estimates, maximum a posteriori (MAP) parameter estimate is used to define the mode of prior mean distribution to improve the performance of MPELR. Experiments on a large vocabulary speech recognition task show that the proposed method can obtain 8.4% relative reduction in word error rate when the amount of data is limited, while retaining the same asymptotic performance as conventional MPELR. When compared with discriminative maximum a posteriori linear regression (DMAPLR), the proposed method shows improvement except for the case of limited adaptation data for supervised adaptation. key words: speaker adaptation (SA), maximum likelihood linear regression (MLLR), maximum a posteriori linear regression (MAPLR), minimum phone error linear regression (MPELR), discriminative maximum a posteriori linear regression (DMAPLR)