Method and system for processing dictated information

At least one non-transitory computer-readable storage medium having encoded thereon computer-executable instructions that, when executed by at least one computer, cause the at least one computer to carry out a method of transcribing speech input from a user to populate a form that includes at least a first field, the method comprising:

• receiving audio of human speech, the human speech comprising audio specifying one or more computer commands and audio corresponding to one or more textual inputs;
• performing automatic speech recognition (ASR), using an ASR engine, on the audio to produce a set of recognition results including the one or more computer commands and a transcription of the one or more textual inputs, wherein performing the ASR on the audio comprises,
• in response to recognizing in the audio a first computer command that is related to a second field available for inclusion in the form, configuring the ASR engine to recognize at least a portion of audio following the first computer command using topic-specific information, the topic-specific information and the second field being related to a same topic, and
• recognizing, using the ASR engine configured with the topic-specific information, the at least the portion of the audio following the first computer command to yield a first textual input;
• populating the form based on at least a part of the set of recognition results, the populating comprising
• in response to detecting in the set the first computer command, adding to the form the second field, and
• populating the second field with the first textual input that appears in the set;
• and
• storing the form in at least one storage medium.

An apparatus comprising:

• at least one processor;
• and
• at least one computer-readable storage medium having encoded thereon executable instructions that, when executed by the at least one processor, cause the at least one processor to carry out a method of transcribing speech input from a user to populate a form that includes at least a first field, the method comprising:
• receiving audio of human speech, the human speech comprising audio specifying one or more computer commands and audio corresponding to one or more textual inputs;
• performing automatic speech recognition (ASR) on the audio, using an ASR engine, to produce a set of recognition results including the one or more computer commands and a transcription of the one or more textual inputs, wherein performing the ASR on the audio comprises,
• in response to recognizing in the audio a first computer command that is related to a second field available for inclusion in the form, configuring the ASR engine to recognize at least a portion of audio following the first computer command using at least one topic-specific language model, the at least one topic-specific language model and the second field being related to a same topic, and
• recognizing, using the ASR engine configured with the at least one topic-specific language model, the at least the portion of the audio following the first computer command to yield a first textual input;
• populating the form based on at least a part of the set of recognition results, the populating comprising
• in response to detecting in the set the first computer command, adding to the form the second field, and
• populating the second field with the first textual input that appears in the set;
• and
• storing the form in at least one storage medium.