HMM-based echo model for noise cancellation avoiding the problem of false triggers

• US 6,606,595 B1
• Filed: 08/31/2000
• Issued: 08/12/2003
• Est. Priority Date: 08/31/2000
• Status: Expired due to Fees

• Current Assignee: Alcatel-Lucent USA, Inc. (Nokia Corporation)
• Original Assignee: Lucent Technologies, Inc. (Nokia Corporation)
• Inventors: Chengalvarayan, Rathinavelu, Setlur, Anand Rangaswamy, Thomson, David Lynn, Ketchum, Richard Harry
• Primary Examiner(s): To, Doris H.
• Assistant Examiner(s): Nolan, Daniel A.
• Application Number:
  US09/652,398
• Time in Patent Office: 1,076 Days
• Field of Search: 704/256, 704/233, 704/228, 704/231, 704/226, 706/45
• US Class Current: 704/256.5
• CPC Class Codes: G10L 15/142 Hidden Markov Models [HMMs]
• CPC Class Codes: G10L 15/20 Speech recognition techniques specially adapted for robustness in adverse environments, e.g. in noise, of stress induced speech G10L21/02 takes precedence
• CPC Class Codes: G10L 2021/02082 the noise being echo, reverberation of the speech

Abstract

An automatic speech recognition system for the condition that an incoming caller'"'"'s speech is quiet and a resulting echo (of a loud playing prompt) can cause the residual (the portion of the echo remaining after even echo cancellation) to be of the magnitude of the incoming speech input. Such loud echoes can falsely trigger the speech recognition system and interfere with the recognition of valid input speech. An echo model has been proven to alleviate this fairly common problem and to be effective in eliminating such false triggering. Further, this automatic speech recognition system enhanced the recognition of valid speech was provided within an existing hidden Markov modeling framework.

Claims

1. A method for preventing a false triggering error from an echo of an audible prompt in an interactive automatic speech recognition system which uses a plurality of hidden Markov models of the system'"'"'s vocabulary with each of the hidden Markov models corresponding to a phrase that is at least one word long, comprising the steps of:

building a hidden Markov model of the audible prompt'"'"'s echo from a plurality of samples of the audible prompt'"'"'s echo;

receiving an input which includes signals that correspond to a caller'"'"'s speech and an echo of the audible prompt of the interactive automatic speech response system; and

using the hidden Markov model of the audible prompt'"'"'s echo along with the plurality of hidden Markov models of the system'"'"'s vocabulary in said automatic speech recognition system to recognize said input when an energy of said echo of the audible prompt is at least the same order of magnitude as the energy of the signals that correspond to the caller'"'"'s speech instead of falsely triggering recognition of one of the plurality of hidden Markov models of the vocabulary.

2. The A method for preventing a false triggering error from an echo of any audible prompt of a plurality of audible prompts in an interactive automatic speech recognition system which uses a plurality of hidden Markov models of the system'"'"'s vocabulary with each of the hidden Markov models corresponding to a phrase that is at least one word long, comprising the steps of:

building hidden Markov models for each of said plurality of audible prompts from a respective plurality of samples of each of the audible prompt'"'"'s echoes;

receiving an input which includes signals that correspond to a caller'"'"'s speech and an echo of one of the plurality of audible prompts of the interactive automatic speech response system; and

using the hidden Markov models of the plurality of audible prompt echoes along with the plurality of hidden Markov models of the system'"'"'s vocabulary in said automatic speech recognition system to recognize said input when an energy of the echo of the audible prompt is at least the same order of magnitude as the energy of the signals that correspond to the caller'"'"'s speech instead of falsely triggering recognition of one of the plurality of hidden Markov models of the vocabulary.