System and Method for Learning Alternate Pronunciations for Speech Recognition

• US 20170032780A1
• Filed: 10/12/2016
• Published: 02/02/2017
• Est. Priority Date: 10/16/2013
• Status: Active Grant

• Current Assignee: Interactive Intelligence Group Incorporated (Genesys Cloud Services Incorporated)
• Original Assignee: Interactive Intelligence Group Incorporated (Genesys Cloud Services Incorporated)
• Inventors: Ge, Zhenhao, Tyagi, Vivek, Ganapathiraju, Aravind, Iyer, Ananth Nagaraja, Randal, Scott Allen, Wyss, Felix Immanuel
• Granted Patent: US 9,767,792 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 40/242 Dictionaries
• CPC Class Codes: G06F 40/40 Processing or translation of natural language natural language analysis G06F40/20; semantic analysis G06F40/30
• CPC Class Codes: G09B 19/04 Speaking with audible presentation of the material to be studied G09B5/04
• CPC Class Codes: G09B 19/06 Foreign languages with audible presentation of material to be studied G09B5/04
• CPC Class Codes: G10L 15/063 Training
• CPC Class Codes: G10L 15/14 using statistical models, e.g. Hidden Markov Models [HMMs] G10L15/18 takes precedence
• CPC Class Codes: G10L 15/187 Phonemic context, e.g. pronunciation rules, phonotactical constraints or phoneme n-grams
• CPC Class Codes: G10L 2015/081 Search algorithms, e.g. Baum-Welch or Viterbi

Abstract

A system and method for learning alternate pronunciations for speech recognition is disclosed. Alternative name pronunciations may be covered, through pronunciation learning, that have not been previously covered in a general pronunciation dictionary. In an embodiment, the detection of phone-level and syllable-level mispronunciations in words and sentences may be based on acoustic models trained by Hidden Markov Models. Mispronunciations may be detected by comparing the likelihood of the potential state of the targeting pronunciation unit with a pre-determined threshold through a series of tests. It is also within the scope of an embodiment to detect accents.

Claims

1. A method for generating candidate pronunciations for a selected word for learning alternative pronunciations of the word in a given language utilizing a grammar-based recognizer in a speech recognition system, the method comprising the steps of:

a. training an acoustic model for use by the grammar-based recognizer on a large speech corpus to distinguish phonemes;

b. constructing a phoneme confusion matrix for application by the grammar-based recognizer to find similar phonemes of mispronounced phonemes in the selected word;

c. constructing a phoneme replacement candidate list of the selected word for each phoneme in a set of speech data containing pronunciations for recognition, using the phoneme confusion matrix;

d. learning, by the grammar-based recognizer, candidate pronunciations of the word, using input from the acoustic model;

e. combining said learned candidate pronunciations with a linguistic dictionary to create a pooled dictionary; and

f. pruning said pooled dictionary to limit the number of learned candidate pronunciations in order to create an improved dictionary.

2. The method of claim 1, wherein the acoustic model in step (a) is trained by one of:

maximum likelihood criterion and discriminative training criterion.

3. The acoustic model of claim 1, wherein said acoustic model in step (a) is based on a Hidden Markov Model and Gaussian Mixture Model.

4. The method of claim 1, wherein the constructing of step (b) comprises merging an acoustic confusion matrix with a linguistic confusion matrix.

5. The acoustic confusion matrix of claim 4, wherein a low value in the acoustic confusion matrix indicates a phoneme is similar to other phonemes and confusable.

6. The linguistic confusion matrix of claim 4, wherein the linguistic confusion matrix comprises a binary matrix comprising the numbers 0 and 1.

7. The linguistic confusion matrix of claim 6, wherein 0 indicates that a phoneme belongs to the same confusion cluster as an other phoneme and the phonemes are confusable.

8. The method of claim 1, wherein the constructing of step (c) further comprises the steps of:

a. selecting a phoneme from the speech data as a target phoneme for analysis and arranging the remaining phonemes based on distance to the target phoneme;

b. applying a statistical clustering algorithm to similarly group the arranged phonemes;

c. constructing the list of phoneme replacement candidates for the target phoneme from the similarly grouped phonemes; and

d. repeating all of the steps for each phoneme in the speech data set.

9. The method of claim 8, wherein the distance between a phoneme and said target phoneme in step (a) represents a confusion value in a phoneme confusion matrix.

10. The phoneme confusion matrix of claim 9, wherein a low value indicates high confusion between a phoneme and a target phoneme.

11. The method of claim 1, wherein the learning of step (d) further comprises:

a. obtaining an original pronunciation for the selected word that has been misrecognized;

b. generating an alternative pronunciation data set for the selected word, wherein an improved new pronunciation is compared with an original pronunciation;

c. performing recognition on the alternative pronunciation data set, the acoustic model, and the set of speech data;

d. determining a best pronunciation from the alternative pronunciation data set; and

e. retaining selected pronunciations from said alternative pronunciation data set, wherein the selected pronunciations are retained to form a learned pronunciation data set.

12. The method of claim 11, wherein the original pronunciation in step (a) is obtained from one of:

a linguistic dictionary and an automatic word-to-phoneme generator.

13. The method of claim 11, wherein generating an alternative pronunciation data set further comprises:

a. placing groups of phonemes in their respective positions; and

b. obtaining all possible phoneme combinations of the phonemes.

14. The method of claim 11, wherein the determination of the best pronunciation comprises basing the determination on pronunciation which results in the highest recognition accuracy.

15. The method of claim 11, wherein the recognition of step (c) is performed using a Viterbi decoding algorithm.

16. The linguistic dictionary of claim 1, wherein the linguistic dictionary comprises a set of pronunciations of common words in a language.

17. The method of claim 11, wherein step (b) further comprises the step of:

determining the size of the alternative pronunciation data set by the mathematical equation;

X=Π

_m=1^M N_m

where M represents a number of phonemes and N, represents a number of phoneme candidates.

18. The method of claim 1, wherein creation of the improved dictionary in step (f) further comprises the steps of:

a. determining a distance from each word to an other word in the linguistic dictionary;

b. creating a subset of similar words for each misrecognized word;

c. performing recognition on the subset of similar words, the acoustic model and the set of speech;

d. identifying a frequency of failure;

e. removing a pronunciation contributing to frequency failure greater than a threshold; and

f. repeating the process for all misrecognized words.

19. The method of claim 18, wherein identifying a frequency of failure further comprises identifying:

incorrect recognitions, a pronunciation associated with an incorrect recognition, and the frequency of failure related to an incorrect recognition.

20. The method of claim 1 further comprising the step of optimizing the efficiency of learning candidate pronunciations, wherein said optimizing comprises one or more of the following:

a. reducing the length of a phoneme replacement candidate list for each phoneme in the original pronunciation of a word that has been mispronounced if a number of candidate pronunciations exceed a threshold; and

b. optimizing a phoneme determination order when obtaining a desired pronunciation for a misrecognized word.

21. The method of claim 20, wherein step (a) further comprises the step of determining the scale of length reduction of a phoneme replacement candidate list with the mathematical equation:

22. The method of claim 20, wherein the phoneme determination order in step (b) continues from a phoneme with the longest phoneme replacement candidate list and continues in the descending order of the length of phoneme replacement candidate list for each phoneme.

23. A method for learning alternative pronunciations for a selected word instance in a given language in a speech recognition system, wherein the speech recognition system comprises at least a grammar-based recognizer, the method comprising the steps of:

a. performing a first test, by the grammar-based recognizer, on the selected word instance to determine a baseline recognition result;

b. performing, by the grammar-based recognizer, hierarchical pronunciation learning on the word instance and selecting a pronunciation that is similar to the word instance; and

c. performing, by the grammar-based recognizer, an other test to determine if the selected pronunciation is recognized as the word instance wherein if the word is recognized, adding the selected pronunciation to a dictionary, otherwise, discarding the selected pronunciation.

24. The method of claim 23, wherein the first test comprises multi-grammar recognition with reference pronunciations.

25. The method of claim 24, wherein the reference pronunciations are scored against the word instance to determine matches.

26. The method of claim 23, wherein the hierarchical pronunciation learning comprises learning alternative pronunciations for the word instance through iterations.

27. The method of claim 23, wherein the other test comprises multi-grammar recognition.

28. A system for language learning of mispronunciation detection for a word in a speech recognition system comprising:

a. a lexicon builder, wherein the lexicon builder is capable of integrating one or more of;

pronunciation dictionaries, spelling-to-pronunciation interpretations, and text normalizations;

b. a speech corpus comprising audio data of pronunciations of the word for recognition;

c. an acoustic model for recognizing pronunciations of the word as phoneme sequences;

d. a word lexicon, wherein the word lexicon provides reference pronunciations of the word;

e. a word grammar, wherein the word grammar specifies words for recognition;

f. a grammar-based recognizer which provides a hypothesized word to a means for scoring, based on input from;

the speech corpus, the acoustic model, the word lexicon, and the word grammar; and

g. a means for scoring which indicates accuracy of the hypothesized word from the grammar-based recognizer, wherein the means for scoring utilizes input from the speech corpus in indicating accuracy.

29. The system of claim 28, wherein the pronunciation dictionaries comprise at least one of:

a learned dictionary, a linguistic dictionary, and prototype dictionary.