Direct Speech-to-speech Translation without Textual Annotation using Bottleneck Features

Open in new window