Domain adaptive speech recognition using artificial intelligence

A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computing device to cause the computing device to:

• generate a set of language data candidates, each language data candidate comprising one or more graphemes, by processing a sequence of phonemes related to user-provided domain-specific input speech data using an artificial intelligence-based data conversion model comprising a neural network model sharing a prediction network from a recurrent neural network transducer;
• determine, for a target pair of one or more phonemes and one or more graphemes, a subset of graphemes from the set of language data candidates;
• train at least one biasing language model using at least a portion of the subset of graphemes;
• generate a first speech recognition output by processing the at least a portion of the subset of graphemes using the at least one biasing language model and an artificial intelligence-based speech recognition model comprising the recurrent neural network transducer, including the prediction network shared by the artificial intelligence-based data conversion model;
• generate a second speech recognition output by replacing at least a portion of the subset of graphemes in the first speech recognition output with at least one of the one or more graphemes from the target pair;
• and
• perform one or more automated actions based at least in part on the second speech recognition output.

A system comprising:

• a memory configured to store program instructions;
• and
• a processor operatively coupled to the memory to execute the program instructions to:
• generate a set of language data candidates, each language data candidate comprising one or more graphemes, by processing a sequence of phonemes related to user-provided domain-specific input speech data using an artificial intelligence-based data conversion model comprising a neural network model sharing a prediction network from a recurrent neural network transducer;
• determine, for a target pair of one or more phonemes and one or more graphemes, a subset of graphemes from the set of language data candidates;
• train at least one biasing language model using at least a portion of the subset of graphemes;
• generate a first speech recognition output by processing the at least a portion of the subset of graphemes using the at least one biasing language model and an artificial intelligence-based speech recognition model comprising the recurrent neural network transducer, including the prediction network shared by the artificial intelligence-based data conversion model;
• generate a second speech recognition output by replacing at least a portion of the subset of graphemes in the first speech recognition output with at least one of the one or more graphemes from the target pair;
• and
• perform one or more automated actions based at least in part on the second speech recognition output.