US 6,725,194 B1Granted

Speech recognition device with text comparing means

AssigneeKoninklijke Philips Electronics N.V.

InventorsHeinrich Bartosik, Walter Müller, Martin Schatz

Life of this patent

Estimating the term…

Loading patent…

Abstract

In a speech recognition device (1) comprising receiving means (36) for receiving voice information (AI) uttered by a speaker and including speech coefficient memory means (38) for storing a speech coefficient indicator (SKI, PRI, SMI, WI) and including speech recognition means (42) which are arranged for recognizing text information (RTI) corresponding to the received voice information (AI) by evaluating the voice information (AI) and the speech coefficient indicator (SKI, PRI, SMI, WI), and including correction means (49) for correcting the recognized text information (RTI) and for producing corrected text information (CTI), text comparing means (52) are provided for comparing the recognized text information (RTI) with the corrected text information (CTI) and for determining at least a correspondence indicator (CI) and the adjusting means (50) are provided for adjusting the stored speech coefficient indicator (SKI, PRI, SMI, WI) by evaluating only one of such text parts (P2) of the corrected text information (CTI) whose correspondence indicator (CI) has a minimum value (MW).

Claims7

Dependent claims

A speech recognition device (1) comprising:

receiving means (36) for receiving voice information (AI) uttered by a speaker and
speech coefficient memory means (38) for storing a speech coefficient indicator (SKI, PRI, SMI, WI) and
speech recognition means (42) which are provided for recognizing text information (RTI) corresponding to the received voice information (AI) by an evaluation of the voice information (AI) and the speech coefficient indicator (SKI, PRI, SMI, WI), and
correction means (49) for correcting the recognized text information (RTI) and for producing corrected text information (CTI), and
adjusting means (50) for adjusting the stored speech coefficient indicator (SKI, PRI, SMI, WI) by evaluation of at least the corrected test information, wherein
text comparing means (42) for comparing the recognized text information (RTI) with the correct text information (CTI) and for defining at least one correspondence indicator (CI) are provided and
in that the adjusting means (50) are provided for automatically adjusting the stored speech coefficient indicator (SKI, PRI, SMI, WI) by programmatically evaluating only one of such text parts (P2) of the corrected text information (CTI) whose correspondence indicator (CI) has a minimum value (MW)), and corrected text information (CTI) that was not associated with the voice information (AI), but was inserted/substituted by a user as correction/substitution that was not an erroneous recognition of text information (RTI) by the speech recognizer, is not used to adjust the stored speech coefficient indicator (SKI, PRI, SMI, WI).

4 more elements

Dependencies are inferred from the claim text.

Facts

Application: 09/610,714Filed 6 Jul 2000
Priority: EP 998902328 Jul 1999
Granted: 20 Apr 2004
Term ends: Not estimatedThe source records don’t support an estimate.

Classification5 codes

G10L 15/075
To the speaker: supervised, i.e. under machine guidance
G10L 15/063
Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice: training
G10L 15/187
Using context dependencies, e.g. language models: phonemic context, e.g. pronunciation rules, phonotactical constraints or phoneme n-grams
G10L 15/197
Grammatical context, e.g. disambiguation of the recognition hypotheses based on word sequence rules: probabilistic grammars, e.g. word n-grams

Family

No related US applications are recorded in the indexed filing history.

Timeline

Citations

Open elsewhere

Building an agent or pulling many patents? Use the API or MCP server instead of scraping pages. Get an API key