Confidence estimation, OOV detection and language ID using phone-to-word transduction and phone-level alignments

White, C.; Zweig, Geoffrey; Burget, Lukáš; Schwarz, Petr; Heřmanský, Hynek

doi:10.1109/icassp.2008.4518552

Cited by 26 publications

(11 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[17,18]) and by adding efficient features (e.g. [6,7,8,9,18]). Second, we will revise our discriminative ETC methods so that they are capable of detecting multiple (consecutive) deletion errors and estimating their number (i.e.…”

Section: Discussionmentioning

confidence: 99%

“…They consist of the features (IDs 1 to 10) used in the recent discriminative model based confidence estimation and OOV detection methods (e.g. [6,7,8,9]) and the WANbased CSID probabilities with some additional features (IDs 11 to 17).…”

Section: Discriminative Error Type Classificationmentioning

confidence: 99%

“…Since ETC is an extension of confidence estimation, we use a discriminative model for ETC with many kinds of features, as with the recent trends in confidence estimation (and out-ofvocabulary word detection), e.g. [6,7,8,9]. Our discriminative ETC receives CSID probabilities provided by a WAN as features, and refines them by exploiting other useful features (i.e.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Recognition rate estimation based on word alignment network and discriminative error type classification

Ogawa

Hori

Nakamura

2012

2012 IEEE Spoken Language Technology Workshop (SLT)

View full text Add to dashboard Cite

Techniques for estimating recognition rates without using reference transcriptions are essential if we are to judge whether or not speech recognition technology is applicable to a new task. This paper proposes two recognition rate estimation methods for continuous speech recognition. The first is an easy-to-use method based on a word alignment network (WAN) obtained from a word confusion network through simple conversion procedures. A WAN contains the correct (C), substitution error (S), insertion error (I) and deletion error (D) probabilities word-by-word for a recognition result. By summing these CSID probabilities individually, the percent correct and word accuracy (WACC) can be estimated without using a reference transcription. The second more advanced method refines the CSID probabilities provided by a WAN based on discriminative error type classification (ETC) and estimates the recognition rates more accurately. In the experiments on the MIT lecture speech corpus, we obtained 0.97 of correlation coefficient between the true WACCs calculated by a scoring tool using reference transcriptions and the WACCs estimated from the discriminative ETC results.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discriminative Error Type Classificationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Recognition rate estimation based on word alignment network and discriminative error type classification

Ogawa

Hori

Nakamura

2012

2012 IEEE Spoken Language Technology Workshop (SLT)

View full text Add to dashboard Cite

show abstract

“…A related area of work is ASR confidence estimation, which seeks to label erroneous words in an ASR hypothesis. [14] uses the comparison of phones in a strong ASR system and a weak ASR system without a language model as features for error detection. Regions where the difference is large indicate a higher likelihood of errors.…”

Section: Related Workmentioning

confidence: 99%

Phonetically-oriented word error alignment for speech recognition error analysis in speech translation

Ruiz

Federico

2015

2015 IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU)

View full text Add to dashboard Cite

We propose a variation to the commonly used Word Error Rate (WER) metric for speech recognition evaluation which incorporates the alignment of phonemes, in the absence of time boundary information. After computing the Levenshtein alignment on words in the reference and hypothesis transcripts, spans of adjacent errors are converted into phonemes with word and syllable boundaries and a phonetic Levenshtein alignment is performed. The phoneme alignment information is used to correct the word alignment labels in each error region. We demonstrate that our Phonetically-Oriented Word Error Rate (POWER) yields similar scores to WER with the added advantages of better word alignments and the ability to capture one-to-many alignments corresponding to homophonic errors in speech recognition hypotheses. These improved alignments allow us to better trace the impact of Levenshtein error types in speech recognition on downstream tasks such as speech translation.

show abstract

“…Numerous works have addressed this aspect of LVCSR in various application scenarios. One scenario is OOV detection, i.e., finding the parts of the utterance that correspond to words unknown to the recognizer [1], [2], [3] where popular approaches include using generalized word or subword filler models, aligning word and subword representations and using confidence measures obtained from various sources. The other scenario is OOV modeling, i.e., the derivation of a lexical representation for the speech component of an unknown word.…”

Section: Introductionmentioning

confidence: 99%

Subword-based automatic lexicon learning for Speech Recognition

Mertens

Seneff

2011

2011 IEEE Workshop on Automatic Speech Recognition &Amp; Understanding

View full text Add to dashboard Cite

We present a framework for learning a pronunciation lexicon for an Automatic Speech Recognition (ASR) system from multiple utterances of the same training words, where the lexical identities of the words are unknown. Instead of only trying to learn pronunciations for known words we go one step further and try to learn both spelling and pronunciation in a joint optimization. Decoding based on linguistically motivated hybrid subword units generates the joint lexical search space, which is reduced to the most appropriate lexical entries based on a set of simple pruning techniques. A cascade of letter and acoustic pruning, followed by re-scoring N -best hypotheses with discriminative decoder statistics resulted in optimal lexical entries in terms of both spelling and pronunciation. Evaluating the framework on English isolated word recognition, we achieve reductions of 7.7% absolute on word error rate and 20.9% absolute on character error rate over baselines that use no pruning.

show abstract

Confidence estimation, OOV detection and language ID using phone-to-word transduction and phone-level alignments

Cited by 26 publications

References 5 publications

Recognition rate estimation based on word alignment network and discriminative error type classification

Recognition rate estimation based on word alignment network and discriminative error type classification

Phonetically-oriented word error alignment for speech recognition error analysis in speech translation

Subword-based automatic lexicon learning for Speech Recognition

Contact Info

Product

Resources

About