Gradual modeling of the Lombard effect by modifying speaker embeddings from a Text-To-Speech model.

Thiago Henrique Gomes Lobato, Magnus Schfer

Browse the full Interspeech paper archive.