Discriminative training of document transcription system

• US 20060074656A1
• Filed: 09/16/2005
• Published: 04/06/2006
• Est. Priority Date: 08/20/2004
• Status: Active Grant

• Current Assignee: Solventum Intellectual Properties Company (Solventum Corp.)
• Original Assignee: Multimodal Technologies Incorporated (3M Company)
• Inventors: Fritsch, Juergen, Yegnanarayanan, Girija, Mathias, Lambert
• Granted Patent: US 8,412,521 B2
• Time in Patent Office: Days
• US Class Current: 704/243
• CPC Class Codes: G06F 40/211 Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
• CPC Class Codes: G06F 40/289 Phrasal analysis, e.g. finite state techniques or chunking
• CPC Class Codes: G06F 40/40 Processing or translation of natural language natural language analysis G06F40/20; semantic analysis G06F40/30
• CPC Class Codes: G10L 15/02 Feature extraction for speech recognition; Selection of recognition unit
• CPC Class Codes: G10L 15/063 Training
• CPC Class Codes: G10L 15/183 using context dependencies, e.g. language models
• CPC Class Codes: G10L 15/193 Formal grammars, e.g. finite state automata, context free grammars or word networks
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 2015/0631 Creating reference templates; Clustering
• CPC Class Codes: G10L 2015/0633 using lexical or orthographic knowledge sources
• CPC Class Codes: G16H 15/00 ICT specially adapted for medical reports, e.g. generation or transmission thereof

Abstract

A system is provided for training an acoustic model for use in speech recognition. In particular, such a system may be used to perform training based on a spoken audio stream and a non-literal transcript of the spoken audio stream. Such a system may identify text in the non-literal transcript which represents concepts having multiple spoken forms. The system may attempt to identify the actual spoken form in the audio stream which produced the corresponding text in the non-literal transcript, and thereby produce a revised transcript which more accurately represents the spoken audio stream. The revised, and more accurate, transcript may be used to train the acoustic model using discriminative training techniques, thereby producing a better acoustic model than that which would be produced using conventional techniques, which perform training based directly on the original non-literal transcript.

Claims

1. In a system including a first document containing at least some information in common with a spoken audio stream, a method comprising steps of:

(A) identifying text in the first document representing a concept having a plurality of spoken forms;

(B) replacing the identified text with a context-free grammar specifying the plurality of spoken forms of the concept to produce a second document;

(C) generating a first language model based on the second document;

(D) using the first language model in a speech recognition process to recognize the spoken audio stream and thereby to produce a third document;

(E) filtering text from the third document by reference to the second document to produce a filtered document in which text filtered from the third document is marked as unreliable; and

(F) using the filtered document and the spoken audio stream to train an acoustic model by performing steps of;

(F) (1) applying a first speech recognition process to the spoken audio stream using a set of base acoustic models and a grammar network based on the filtered document to produce a first set of recognition structures;

(F) (2) applying a second speech recognition process to the spoken audio stream using the set of base acoustic models and a second language model to produce a second set of recognition structures; and

(F) (3) performing discriminative training of the acoustic model using the first set of recognition structures, the second set of recognition structures, the filtered document, and only those portions of the spoken audio stream corresponding to text not marked as unreliable in the filtered document.

2. The method of claim 1, further comprising a step of:

(F) (4) prior to (F) (1), training the set of base acoustic models using the spoken audio stream and the filtered document.

3. The method of claim 2, wherein step (F) (4) comprises training the set of base acoustic models using maximum likelihood optimization training.

4. The method of claim 1, wherein the discriminative training comprises maximum mutual information estimation training, wherein the first set of recognition structures comprises a “

correct”

lattice, and wherein the second set of recognition structures comprises a “

general”

lattice.

5. A system comprising:

a first document containing at least some information in common with a spoken audio stream;

means for identifying text in the first document representing a concept having a plurality of spoken forms;

means for replacing the identified text with a context-free grammar specifying the plurality of spoken forms of the concept to produce a second document;

means for generating a first language model based on the second document;

means for using the first language model in a speech recognition process to recognize the spoken audio stream and thereby to produce a third document;

means for filtering text from the third document by reference to the second document to produce a filtered document in which text filtered from the third document is marked as unreliable; and

means for using the filtered document and the spoken audio stream to train an acoustic model, comprising;

means for applying a first speech recognition process to the spoken audio stream using a set of base acoustic models and a grammar network based on the filtered document to produce a first set of recognition structures;

means for applying a second speech recognition process to the spoken audio stream using the set of base acoustic models and a second language model to produce a second set of recognition structures; and

means for performing discriminative training of the acoustic model using the first set of recognition structures, the second set of recognition structures, the filtered document, and only those portions of the spoken audio stream corresponding to text not marked as unreliable in the filtered document.

6. The system of claim 5, further comprising:

means for training the set of base acoustic models using the spoken audio stream and the filtered document.

7. The system of claim 6, wherein the means for training the set of base acoustic models comprises means for training the set of base acoustic models using maximum likelihood optimization training.

8. The system of claim 5, wherein the discriminative training comprises maximum mutual information estimation training, wherein the first set of recognition structures comprises a “

correct”

lattice, and wherein the second set of recognition structures comprises a “

general”

lattice.

9. A method comprising steps of:

(A) identifying a normalized document containing at least some information in common with a spoken audio stream, the normalized document including a context-free grammar specifying a plurality of spoken forms of a concept;

(B) identifying a language model based on the normalized document;

(C) using the language model in a speech recognition process to recognize the spoken audio stream and thereby to produce a second document;

(D) filtering text from the second document by reference to the normalized document to produce a filtered document in which text filtered from the second document is marked as unreliable; and

(E) using the filtered document and the spoken audio stream to train an acoustic model by performing steps of;

(E) (1) applying a first speech recognition process to the spoken audio stream using a set of base acoustic models and a grammar network based on the filtered document to produce a first set of recognition structures;

(E) (2) applying a second speech recognition process to the spoken audio stream using the set of base acoustic models and a second language model to produce a second set of recognition structures; and

(E) (3) performing discriminative training of the acoustic model using the first set of recognition structures, the second set of recognition structures, the filtered document, and only those portions of the spoken audio stream corresponding to text not marked as unreliable in the filtered document.

10. The method of claim 9, further comprising a step of:

(E) (4) prior to (E) (1), training the set of base acoustic models using the spoken audio stream and the filtered document.

11. The method of claim 10, wherein step (E) (4) comprises training the set of base acoustic models using maximum likelihood optimization training.

12. The method of claim 9, wherein the discriminative training comprises maximum mutual information estimation training, wherein the first set of recognition structures comprises a “

correct”

lattice, and wherein the second set of recognition structures comprises a “

general”

lattice.

13. A system comprising:

means for identifying a normalized document containing at least some information in common with a spoken audio stream, the normalized document including a context-free grammar specifying a plurality of spoken forms of a concept;

means for identifying a language model based on the normalized document;

means for using the language model in a speech recognition process to recognize the spoken audio stream and thereby to produce a second document;

means for filtering text from the second document by reference to the normalized document to produce a filtered document; and

means for using the filtered document and the spoken audio stream to train an acoustic model, comprising;

means for applying a first speech recognition process to the spoken audio stream using a set of base acoustic models and a grammar network based on the filtered document to produce a first set of recognition structures;

means for applying a second speech recognition process to the spoken audio stream using the set of base acoustic models and a second language model to produce a second set of recognition structures; and

means for performing discriminative training of the acoustic model using the first set of recognition structures, the second set of recognition structures, the filtered document, and only those portions of the spoken audio stream corresponding to text not marked as unreliable in the filtered document.

14. The system of claim 13, further comprising:

means for training the set of base acoustic models using the spoken audio stream and the filtered document.

15. The system of claim 14, wherein the means for training the set of base acoustic models comprises means for training the set of base acoustic models using maximum likelihood optimization training.

16. The system of claim 13, wherein the discriminative training comprises maximum mutual information estimation training, wherein the first set of recognition structures comprises a “

correct”

lattice, and wherein the second set of recognition structures comprises a “

general”

lattice.