The IBM expressive text-to-speech synthesis system for American English

A review of paralinguistic information processing for natural speech communication

“…The US-based ESS requires a large scale corpus of emotional speech, and emotional property of synthetic speech is colored by acoustic correlates represented by inventories in a corpus [2,[47][48][49][50][51]. The quality of ESS is dependent on the corpus and how to select inventories from the corpus.…”

“…This approach is simple but it needs large costs for constructing corpora. Pitrelli et al proposed another approach of the USbased ESS by building a corpus that mixed several emotions and by introducing the emotion as a feature for inventory selection [49]. Moriyama et al proposed an idea for representing relationship among F 0 , energy, and duration using PCA on a subspace in ESS for Japanese words [50,51].…”

Frequency Modulation Technique for Prosodic Modification

“…1 shows a set of warping functions, depending on the values of ζ. It is clear that λ ∈ [1,2] can be mapped to multiple F0 values in [f0 l , f0 h ] when ζ is altered. This forms the basis of frequency modulation in that λ and ζ can be used to represent the observed F0 contours and the adjusting proportions, respectively.…”

“…Furthermore, significant progress has been made in corpus-based unit concatenative synthesis technology [2] [3]. These two things have led to an improvement in voice quality of synthetic speech, which in turn has led to it becoming more common.…”

“…In particular, reading-style prosody is far from satisfactory in most situations that involve human-machine dialogs or machine-mediated human-human dialogs. Therefore, there has been a lot of research on expressive and conversational speech synthesis so as to improve expressiveness of synthetic speech [2] [5] [6] [7].…”