Conversion of neutral speech to storytelling style speech

Abstract

Speech is the most basic and widely used method for communication. There is a growing need for an expressive speech synthesis especially, when human want to communicate with robots and computers. In this paper, prosody rule-sets are designed to convert neutral to storytelling style speech for Hindi language. In order to generate a storyteller speech from a neutral speech, modification in various prosodic parameters such as pitch, intensity, duration, tempo and pauses are considered. For each of the above mentioned prosodic parameters rules are developed separately for story-specific emotions (such as sad, anger, fear, surprise and neutral). Theses rules are designed by performing analysis on stories collected from a professional storyteller. In this work, modification are done both at phrase and sentencelevel. These rules are derived for both male and female speakers. Subjective tests are conducted to evaluate the quality of the generated storytelling style speech. Also, influence of speaker characteristics on neutral to story speech conversion are analysed.

The first Malay language storytelling text-to-speech (TTS) corpus for humanoid robot storytellers

“…We began our study by reviewing related work on storytelling speech corpus in different languages such as Bengali [10], Hindi [12], Dutch [13], English [3,14], French [15], Slovak [16] and Spanish [17]. Summary of the review is presented in Table 1 and briefly described accordingly.…”

“…Other researchers that also employed professional speakers as their storytellers are [15], [14]. On the contrary, [10,[12][13] engaged professional artists and actors for their collections of storytelling audio speeches. [10] to record the same stories in a neutral style like news reading speech.…”

“…In [12], the children stories are taken from Internet collections and story books such as Panchatantra and Akbar Birbal.…”

Age Recognition Based on GMM and DBN with Using Wavelet Packet Mel-Frequency Cepstrum Coefficient

“…It was realized early that speakers of different age groups have different speech features [2], Nevertheless, the systematic research on the correlation between different age groups' speech including the speaker age recognition and age speech conversion mainly arose from the recent personalize intelligent human-computer interaction system and expressive speech synthesis needs. The most attractive approach was to apply a phonetically-aware deep neural network to achieve speaker recognition [3].…”

Rule-Based Storytelling Text-to-Speech (TTS) Synthesis

“…Therefore, to formulate this risefall pattern of pitch contour, non-linear functions are explored. In this paper, the pitch contour patterns are formulated using sine functions [18]. The formulation multiplies all pitch values within a relevant time domain [t 1 , t 2 ] for a given modification factor which is governed by the variable α .…”