Detail publikace

Recovery from Model Inconsistency in Multilingual Speech Recognition

Originální název

Recovery from Model Inconsistency in Multilingual Speech Recognition

Anglický název

Recovery from Model Inconsistency in Multilingual Speech Recognition

Jazyk

en

Originální abstrakt

Current ASR has difficulties in handling unexpected words that are typically replaced by acoustically acceptable high prior probability words. Identifying parts of the message where such a replacement could have happened may allow for corrective strategies.

We aim to develop data-guided techniques that would yield unconstrained estimates of posterior probabilities of sub-word classes employed in the stochastic model solely from the acoustic evidence, i.e. without use of higher level language constraints.

These posterior probabilities then could be compared with the constrained estimates of posterior probabilities derived with the constraints implied by the underlying stochastic model. Parts of the message where any significant mismatch between these two probability distributions is found should be re-examined and corrective strategies applied.

This may allow for development of systems that are able to indicate when they "do not know" and eventually may be able to "learn-as-you-go" in applications encountering new situations and new languages.

During the 2007 Summer Workshop we intend to focus on detection and description of out-of-vocabulary and mispronounced words in the 6 language Call-home database. Additionally, in order to describe the suspect parts of the message, we will work on language-independent recognizer of speech sounds that could be applied for phonetic transcription of identified suspect parts of the recognized message.

Anglický abstrakt

Current ASR has difficulties in handling unexpected words that are typically replaced by acoustically acceptable high prior probability words. Identifying parts of the message where such a replacement could have happened may allow for corrective strategies.

We aim to develop data-guided techniques that would yield unconstrained estimates of posterior probabilities of sub-word classes employed in the stochastic model solely from the acoustic evidence, i.e. without use of higher level language constraints.

These posterior probabilities then could be compared with the constrained estimates of posterior probabilities derived with the constraints implied by the underlying stochastic model. Parts of the message where any significant mismatch between these two probability distributions is found should be re-examined and corrective strategies applied.

This may allow for development of systems that are able to indicate when they "do not know" and eventually may be able to "learn-as-you-go" in applications encountering new situations and new languages.

During the 2007 Summer Workshop we intend to focus on detection and description of out-of-vocabulary and mispronounced words in the 6 language Call-home database. Additionally, in order to describe the suspect parts of the message, we will work on language-independent recognizer of speech sounds that could be applied for phonetic transcription of identified suspect parts of the recognized message.

BibTex


@techreport{BUT58266,
  author="Hynek {Heřmanský} and Lukáš {Burget} and Petr {Schwarz} and Pavel {Matějka} and Mirko {Hannemann} and Ariya {Rastrow} and Christopher {White} and Sanjeev {Khudanpur} and Jan {Černocký}",
  title="Recovery from Model Inconsistency in Multilingual Speech Recognition",
  annote="Current ASR has difficulties in handling unexpected words that are typically replaced by acoustically acceptable high prior probability words. Identifying parts of the message where such a replacement could have happened may allow for corrective strategies.

We aim to develop data-guided techniques that would yield unconstrained estimates of posterior probabilities of sub-word classes employed in the stochastic model solely from the acoustic evidence, i.e. without use of higher level language constraints.

These posterior probabilities then could be compared with the constrained estimates of posterior probabilities derived with the constraints implied by the underlying stochastic model. Parts of the message where any significant mismatch between these two probability distributions is found should be re-examined and corrective strategies applied.

This may allow for development of systems that are able to indicate when they "do not know" and eventually may be able to "learn-as-you-go" in applications encountering new situations and new languages.

During the 2007 Summer Workshop we intend to focus on detection and description of out-of-vocabulary and mispronounced words in the 6 language Call-home database. Additionally, in order to describe the suspect parts of the message, we will work on language-independent recognizer of speech sounds that could be applied for phonetic transcription of identified suspect parts of the recognized message.", address="Johns Hopkins University", chapter="58266", institution="Johns Hopkins University", year="2007", month="december", pages="0--0", publisher="Johns Hopkins University", type="report" }