← Back to all use cases

Video coming soon

The avatar reads the first act of the Aozora Bunko TEI of “Mittsu no Takara” following a script JSON (json=) that carries emotion, mouth shapes and body motion for each line. Script lines are aligned to the speech in the text, so the line being read is highlighted. No audio files are included, so the voice is the browser’s speech synthesis (with a different pitch per speaker); the expressions and motion come from the script.