Objective methods for evaluating synthetic intonation
نویسندگان
چکیده
This paper describes the development and evaluation of objective methods for testing synthetic intonation. While subjective methods are available for assessing the quality of synthetic intonation, such tests consume time and resources, and are not useful for day-to-day model development. Therefore, objective measures of F0 modelling are necessary. Currently, objective evaluation of synthetic intonation involves the use of Root Mean Squared Error and Correlation. However, it is unclear how large an improvement in either score must be before it is reflected perceptually. It is also unclear how detailed an analysis these metrics provide. Therefore, two other metrics are to be tested, both of which are similar to a basic RMSE measurement. All of the evaluation results are compared to a perceptual study in order to determine how the objective measures relate to perceived differences in the contours.
منابع مشابه
Subjective Evaluation of Synthetic Intonation
This paper describes a method for evaluating the quality of synthetic intonation using subjective techniques. This perceptual method of assessing intonation, not only evaluates the quality of synthetic intonation, but also allows us to compare different models of intonation to know which one is the most natural from a perceptual point of view. This procedure has been used to assess the quality ...
متن کاملExtended ratio edge detector for despeckled SAR image evaluation
Synthetic aperture radar (SAR) images due to the usage of coherent imaging systems are affected by speckle. So lots of despeckling filters have been introduced up to now to suppress the speckle. Hence, objective and subjective evaluation of the denoised SAR images becomes a necessity. Thereby lots of objective evaluating estimators are introduced to evaluate the performance of despeckling filte...
متن کاملIntonation modeling of Mandarin Chinese using a superpositional approach
The intonation model is an important component in text-tospeech systems to obtain natural and expressive speech synthesis. In this paper we propose a superpositional model for Mandarin Chinese. The intonation model is composed of the syllable and the phrase component. The parameters of the model are estimated using JEMA, a training approach with many advantages related to robustness and precisi...
متن کاملAutomatic synthesis of natural-sounding intonation for text-to-speech conversion in dutch
A set of rules is proposed for the automatic synthesis of natural-sounding intonation as part of speech synthesis in Dutch from unrestricted text. Results of a formal perceptual evaluation show that the synthetic intonation is judged to be as natural as human intonation for isolated utterances; for texts, additional provisions are required to model contributions of text structure. It is suggest...
متن کاملThe importance of segmental duration and f0 for generating more natural intonation in synthetic speech
This dissertation presents the importance of diphones’ duration and f0 information in generating more natural intonation in unit selection speech synthesis. The results showed that diphones’ duration or f0 information was highly correlated to one another due to the prosodic properties inherited from the recorded human speech. Also only raising the importance of duration and f0 information large...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 1999