Speaker adaptation for attention-based encoder-decoder

A computer-implemented method comprising:

• training a first attention-based encoder-decoder model for automatic speech recognition, the trained speaker-independent attention-based encoder-decoder model associated with a first output distribution;
• adapting the first attention-based encoder-decoder model for automatic speech recognition of speech of a target speaker, and to maintain a similarity between the first output distribution and a second output distribution of the adapted attention-based encoder-decoder model;
• and
• performing automatic speech recognition using the adapted speaker-independent attention-based encoder-decoder model.

A non-transient, computer-readable medium storing program code to be executed by a processing unit to provide:

• a first attention-based encoder-decoder model to classify output tokens based on input speech frames, the first attention-based encoder-decoder model associated with a first output distribution;
• and
• a second attention-based encoder-decoder model to classify output tokens based on input speech frames, the second attention-based encoder-decoder model being associated with a second output distribution,
• wherein the second attention-based encoder-decoder model is trained to classify output tokens based on input speech frames of a target speaker, and simultaneously trained to maintain a similarity between the first output distribution and a second output distribution of the second attention-based encoder-decoder model.