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

摘要

This paper describes the process undertaken and criteria considered in acquiring a storytellingspeech corpus of Malay language towards the development of humanoid storyteller. Thespeech corpus contains 464 speech sentences, 4,656 words and 9,584 syllables. Threechildren’s short stories were recorded by 3 female storytellers, 1 male professional speaker, 2female speakers and 2 male speakers. The equipment specifications, recording procedures andspeech annotations are described in detail in accordance to baseline work. The stories wererecorded in two speaking styles that are neutral and storytelling speaking style. The firstMalay language storytelling corpus is not only necessary for the development of a storytellingtext-to-speech (TTS) synthesis. It is also detrimental for natural language processing andspeech recognition of Malay language, an under-resourced languageKeywords: storytelling speech corpus; humanoid storyteller; storytelling TTS; Malaylanguage.