Assessment of Standard Arabic Acoustic Phonetic Decoder. Djoudi, M. In Proceedings of the 13th National Computer Conference. Riyadh, Kingdom of Saudi Arabia, 1992. Paper abstract bibtex Within the framework of the automatic recognition of continuous speech, we have developed SAPHA system : an acoustic phonetic decoder of standard Arabic. In This article, we present a first assessment of the system. SAPHA makes possible the analytic recognition of phonemes in continuous speech for multispeakers. Combining other linguistic information (lexicon, syntax, semantic and pragmatic), the system can be considered as an important step towards a system for oral dialog between man and machine. It also acts as a module of a dictation machine. The recognition module is realized as form of an expert system based on production rules. Knowledge is acquired after a phonetic study of Arabic carried out on DJOUMA corpus, which is composed of 50 sentences pronounced by 11 speakers (7 males and 4 Females) . This study also allows us to adopt a strategy for signal segmentation into large phonetic classes and determine the distinctive values of parameters used during recognition. In the same way, the manual labelling of the corpus sentences allows us to test the performances of the system
@inproceedings{djoudi_assessment_1992,
title = {Assessment of {Standard} {Arabic} {Acoustic} {Phonetic} {Decoder}},
copyright = {All rights reserved},
url = {http://djoudi.online.fr/publications/files/1992_11_28_13NCC_Riyadh_Assessment.pdf},
abstract = {Within the framework of the automatic recognition of continuous speech, we have developed SAPHA system : an acoustic phonetic decoder of standard Arabic. In This article, we present a first assessment of the system. SAPHA makes possible the analytic recognition of phonemes in continuous speech for multispeakers. Combining other linguistic information (lexicon, syntax, semantic and pragmatic), the system can be considered as an important step towards a system for oral dialog between man and machine. It also acts as a module of a dictation machine. The recognition module is realized as form of an expert system based on production rules. Knowledge is acquired after a phonetic study of Arabic carried out on DJOUMA corpus, which is composed of 50 sentences pronounced by 11 speakers (7 males and 4 Females) . This study also allows us to adopt a strategy for signal segmentation into large phonetic classes and determine the distinctive values of parameters used during recognition. In the same way, the manual labelling of the corpus sentences allows us to test the performances of the system},
booktitle = {Proceedings of the 13th {National} {Computer} {Conference}. {Riyadh}, {Kingdom} of {Saudi} {Arabia}},
author = {Djoudi, Mahieddine},
year = {1992},
keywords = {Speech Recognition},
}
Downloads: 0
{"_id":"Kkd8nACBq5CiKk69G","bibbaseid":"djoudi-assessmentofstandardarabicacousticphoneticdecoder-1992","downloads":0,"creationDate":"2017-11-27T23:02:16.136Z","title":"Assessment of Standard Arabic Acoustic Phonetic Decoder","author_short":["Djoudi, M."],"year":1992,"bibtype":"inproceedings","biburl":"https://bibbase.org/zotero/mdjoudi","bibdata":{"bibtype":"inproceedings","type":"inproceedings","title":"Assessment of Standard Arabic Acoustic Phonetic Decoder","copyright":"All rights reserved","url":"http://djoudi.online.fr/publications/files/1992_11_28_13NCC_Riyadh_Assessment.pdf","abstract":"Within the framework of the automatic recognition of continuous speech, we have developed SAPHA system : an acoustic phonetic decoder of standard Arabic. In This article, we present a first assessment of the system. SAPHA makes possible the analytic recognition of phonemes in continuous speech for multispeakers. Combining other linguistic information (lexicon, syntax, semantic and pragmatic), the system can be considered as an important step towards a system for oral dialog between man and machine. It also acts as a module of a dictation machine. The recognition module is realized as form of an expert system based on production rules. Knowledge is acquired after a phonetic study of Arabic carried out on DJOUMA corpus, which is composed of 50 sentences pronounced by 11 speakers (7 males and 4 Females) . This study also allows us to adopt a strategy for signal segmentation into large phonetic classes and determine the distinctive values of parameters used during recognition. In the same way, the manual labelling of the corpus sentences allows us to test the performances of the system","booktitle":"Proceedings of the 13th National Computer Conference. Riyadh, Kingdom of Saudi Arabia","author":[{"propositions":[],"lastnames":["Djoudi"],"firstnames":["Mahieddine"],"suffixes":[]}],"year":"1992","keywords":"Speech Recognition","bibtex":"@inproceedings{djoudi_assessment_1992,\n\ttitle = {Assessment of {Standard} {Arabic} {Acoustic} {Phonetic} {Decoder}},\n\tcopyright = {All rights reserved},\n\turl = {http://djoudi.online.fr/publications/files/1992_11_28_13NCC_Riyadh_Assessment.pdf},\n\tabstract = {Within the framework of the automatic recognition of continuous speech, we have developed SAPHA system : an acoustic phonetic decoder of standard Arabic. In This article, we present a first assessment of the system. SAPHA makes possible the analytic recognition of phonemes in continuous speech for multispeakers. Combining other linguistic information (lexicon, syntax, semantic and pragmatic), the system can be considered as an important step towards a system for oral dialog between man and machine. It also acts as a module of a dictation machine. The recognition module is realized as form of an expert system based on production rules. Knowledge is acquired after a phonetic study of Arabic carried out on DJOUMA corpus, which is composed of 50 sentences pronounced by 11 speakers (7 males and 4 Females) . This study also allows us to adopt a strategy for signal segmentation into large phonetic classes and determine the distinctive values of parameters used during recognition. In the same way, the manual labelling of the corpus sentences allows us to test the performances of the system},\n\tbooktitle = {Proceedings of the 13th {National} {Computer} {Conference}. {Riyadh}, {Kingdom} of {Saudi} {Arabia}},\n\tauthor = {Djoudi, Mahieddine},\n\tyear = {1992},\n\tkeywords = {Speech Recognition},\n}\n\n\n\n","author_short":["Djoudi, M."],"key":"djoudi_assessment_1992","id":"djoudi_assessment_1992","bibbaseid":"djoudi-assessmentofstandardarabicacousticphoneticdecoder-1992","role":"author","urls":{"Paper":"http://djoudi.online.fr/publications/files/1992_11_28_13NCC_Riyadh_Assessment.pdf"},"keyword":["Speech Recognition"],"metadata":{"authorlinks":{"djoudi, m":"https://bibbase.org/show?bib=https%3A%2F%2Fbibbase.org%2Fzotero%2Fmdjoudi&msg=embed","djoudi, m":"http://djoudi.online.fr/newsrss.php"}},"downloads":0},"search_terms":["assessment","standard","arabic","acoustic","phonetic","decoder","djoudi"],"keywords":["speech recognition"],"authorIDs":["HRKdyeSzR3o7eMNw7","Y5ExTeNBYWDkytzSJ"],"dataSources":["rh94BkjSH73SyjZsR","ya2CyA73rpZseyrZ8","iepCnBefzQg3Bw5fo","GLMs4nYirTtANpSJL","2252seNhipfTmjEBQ"]}