Method for driving a dialog system

• US 20070150287A1
• Filed: 07/22/2004
• Published: 06/28/2007
• Est. Priority Date: 08/01/2003
• Status: Abandoned Application

• Current Assignee: Koninklijke Philips Electronics N.V. (Koninklijke Philips N.V.)
• Original Assignee: Koninklijke Philips Electronics N.V. (Koninklijke Philips N.V.)
• Inventors: Portele, Thomas, Thiele, Frank
• Application Number:
  US10/566,512
• Publication Number:
  US 20070150287A1
• Time in Patent Office: Days
• US Class Current: 704/275
• CPC Class Codes: G10L 15/1822 Parsing for meaning understanding
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/228 of application context
• CPC Class Codes: G10L 2025/786 Adaptive threshold
• CPC Class Codes: G10L 21/0208 Noise filtering

Abstract

The invention describes a method for driving a dialog system (1) comprising an audio interface (11) for processing audio signals (3,6). The method deduces characteristics (2) of an expected audio input signal (3) and generates audio interface control parameters (4) according to these characteristics (2). The behaviour of the audio interface (11) is optimised based on the audio interface control parameters (4). Moreover the invention describes a dialog system (1) comprising an audio interface (11), a dialog control unit (12), a predictor module (13) for deducing characteristics (2) of an expected audio input signal (3), and an audio optimiser (14) for optimising the behaviour of the audio interface (11) by generating audio input control parameters (4) based on the characteristics (2).

Claims

1. A method for driving a dialog system (1) comprising an audio interface (11) for processing audio signals (3,6) wherein characteristics (2) of an expected audio input signal (3) are deduced, audio interface control parameters (4) are generated according to these characteristics (2), behaviour of the audio interface (11) is optimised based on the audio interface control parameters (4).

2. The method according to claim 1, wherein characteristics (2) are deduced from current and/or prior input data.

3. The method according to claim 2, wherein characteristics (2) are deduced from a semantic analysis of the speech content (5) of the input audio signal (3);

4. The method according to claim 2, wherein characteristics (2) are deduced from determined environmental conditions data.

5. The method according to claim 1, wherein characteristics (2) are deduced from an expected response to a current prompt (6) of the dialog system (1).

6. The method according to claim 1, wherein the control parameters (4) comprise recording and/or processing parameters for an audio driver (9) of the audio interface (11).

7. The method according to claim 1, wherein the control parameters (4) comprise threshold parameters for an audio module (10) of the audio interface (11).

8. A dialog system (1) comprising an audio interface (11), a dialog control unit (12), a predictor module (13) for deducing characteristics (2) of an expected audio input signal (3), an audio optimiser (14) for optimising the behaviour of the audio interface (11) by generating audio input control parameters (4) based on the characteristics (2).

9. The dialog system (1) according to claim 8, wherein the audio interface (11) consists of audio hardware (8) and/or an audio driver (9) and/or an audio module (10).