Convolutional neural network with phonetic attention for speaker verification

A method comprising:

• determining, for each of a plurality of speech frames associated with an acoustic feature, a phonetic feature based on the associated acoustic feature;
• generating one or more two-dimensional feature maps based on the plurality of phonetic features;
• inputting the one or more two-dimensional feature maps to a trained neural network to generate a plurality of speaker embeddings;
• and
• aggregating the plurality of speaker embeddings into a speaker embedding based on respective weights determined for each of the plurality of speaker embeddings,
• wherein the speaker embedding is associated with an identity of the speaker.

A hardware storage device storing program code to be executed by a processing unit to provide:

• a feature extractor to determine a plurality of multi-dimensional phonetic features based on a plurality of multi-dimensional acoustic features;
• a deep neural network to generate a plurality of two-dimensional feature maps based on the plurality of multi-dimensional phonetic features;
• a convolutional neural network to receive the plurality of two-dimensional feature maps and the plurality of multi-dimensional acoustic features and to generate a plurality of speaker embeddings based on the plurality of two-dimensional feature maps and the plurality of multi-dimensional acoustic features;
• and
• an attentive pooling network to determine respective weights for each of the plurality of speaker embeddings and to aggregate the plurality of speaker embeddings into a first speaker embedding based on the respective weights,
• wherein the speaker embedding is associated with an identity of the speaker.