Multi-state barge-in models for spoken dialog systems

The method of applying a multi-state barge-in acoustic model in a spoken dialog system, the method comprising:

• presenting a prompt to a user from the spoken dialog system;
• receiving an audio speech input from the user during the presentation of the prompt;
• accumulating the audio speech input from the user;
• applying a non-speech component having at least two one-state Hidden Markov Models (HMMs) to the audio speech input from the user;
• applying a speech component having at least five three-state HMMs to the audio speech input from the user, wherein each of the five three-state HMMs represents a different phonetic category;
• determining whether the audio speech input is a barge-in-speech input from the user;
• and
• if the audio speech input is determined to be the barge-in-speech input from the user, terminating the presentation of the prompt.

A spoken dialog system that applies a multi-state barge-in acoustic model, the system comprising:

• a module configured to present prompt to a user;
• a module configured to receive an audio speech input from the user during a presentation prompt;
• a module configured to accumulate the audio speech input from the user;
• a module configured to apply a non-speech component having at least two one-state Hidden Markov Models (HMMs) and two three-state left-to-right HMMs;
• and
• a module configured to apply a non-speech component having at least two one-state Hidden Markov Models (HMMs) to the audio speech input from the user;
• a module configured to apply a speech component having at least five three-state HMMs to the audio speech input from the user, wherein each of the five three-state HMMs represents a different phonetic category;
• a module configured to determine whether the audio speech input is a barge-in-speech input from the user;
• and
• a module configured, if the audio speech input is determined to be the barge-in-speech input from the user, to terminate the presentation of the prompt.