Corpus-based speech synthesis based on segment recombination

摘要

A system and method generate synthesized speech through concatenation of speech segments that are derived from a large prosodically-rich corpus of speech segments including using an additional dictionary of speech segment identifier sequences.