Tone correctness improvement in speaker dependent HMM-based Thai speech synthesis

Problem statement:The expressive speech of Thai had been studied for a short period of time. An important feature of speech was fundamental frequency (F0) which defines the human speech prosody. It could be used to distinguish the difference between several types of expressive speech. The environmental noises affect the F0 contour for Thai dialects as concluded in the previous study. The study prosodic information of Thai speech with various speaking styles and several types of noises had not been conducted. Approach: Four different types of speaking styles were used; meanwhile four types of environmental noises were recorded with different levels of power. They were subsequently mixed together. The F0 contours from different types of speaking styles, different types of noises and different levels of noises were extracted. The Root Mean Square Error (RMSE) between the F0 contour of clean speech and the noise-corrupted speech was calculated. Results: In the experiments, four types of noises were included train, factory, car and air conditioner. Each type of speaking style included 10 samples of 10 utterances of male and female speech. Five levels of noises were varied from 0-20 dB compared with the clean speech. It could be notified that the effects of distinguishing types of noises were different. Four different types of speaking styles were also caused the differences in RMSEs. Conclusion: The recorded noises deteriorate the F0 contours for all types of speaking styles in Thai.

Section: Resultsmentioning

confidence: 99%

Effects of Environmental Noises on Fundamental Frequency Contours of Thai Expressive Speech

N.¹

2012

“…Before mixing noises with the clean speech, the noise volume or power are adjusted in several exact levels. As for the level variation of noises, the levels of each type of noise are varied from 0, 5, 10, 15, 20 dB, respectively (Chomphan and Kobayashi, 2008;2009) …”

Section: F0 Extractionmentioning

confidence: 99%

Effects of Noises on Fundamental Frequency Extraction Using Cepstral Analysis for Thai Dialects

Chomphan¹

2012

Problem statement:The fundamental frequency (F0) of the human speech corresponds to the vibration frequency of the human vocal chords. To extract the F0 from a speech utterance, one approach is based on the Cepstral analysis. In Thai, there are four main dialects spoken by Thai people residing in four core region including central, north, northeast and south regions. Environmental noises are also playing an important role in corrupting the speech quality. It is needed to study of effects of noises on F0 extraction using the Cepstral analysis for Thai dialects. Approach: The Cepstral analysis is performed and some coefficients are used to determine the corresponding F0 values. Four types of environmental noises are simulated with different levels of power. The differences among the extracted F0 from clean speech and the extracted F0 from noisecorrupted speech are calculated in Root Mean Square (RMS) errors. Results: The selected noises are train, factory, car and air conditioner. Five levels of each type of noise vary from 0-20 dB. From the experimental results, it has been noticed that the effects of noises are different. The lowest effect is of air conditioner, meanwhile the noise level of 0 dB is of the highest effect. Conclusion: By using the Cepstral analysis, F0 values can be extracted from the noise-corrupted speech with different level of effects depending on the type and level of noises.

“…By appling the Fujisaki's model, the related parameters are extracted from the speech corpus, utterance by utterance. Therafter the output parameters are calculated are systematically analyzed (Chomphan and Kobayashi, 2008;2009;Seresangtakul and Takara, 2003).…”

Section: Introductionmentioning

confidence: 99%

Fujisakiâs Model of Thaiâs Fundamental Frequency Contours with Environmental Noises

Edgar¹

2012

Problem statement: An important human speech feature is the fundamental frequency (F0) contour which represents the speech prosody. It indicates the naturalness and intelligibility of the speech. Modeling of fundamental frequency contour was an essential procedure in the natural speech processing. In speech communication, environmental noise plays an essential role in damaging the digital communication quality. The study of effects of noises on modeling of F0 contour for standard Thai is conducted. Approach: The selected modeling technique in this study was adapted from Fujisaki's model, because of its achievement in modeling of various Thai speech units. Four types of environmental noises were recorded for different levels of power. This study was proposed an analysis of some parameters of modeling of Thai speech prosody for two genders and four types of noises. The derived Fujisaki's model was covered seven parameters including baseline frequency, the numbers of phrase commands and tone commands, phrase command and tone command durations, amplitudes of phrase command and tone command. Results: In the experimental results, the standard Thai of 2 samples of 5 sentences with 5 males and 5 females was used. Four types of noises include train, factory, car and air conditioner. Five levels of each type of noise were varied from 0-20 dB. The results were showing that the different noises give the distinguished effects for most of the proposed model parameters. Conclusion: The results confirm that the effects of four types of noises are significantly different. It can be seen that the environmental noises deteriorate the model parameters empirically.