Word recognition in continuous speech using linear prediction analysis:
A promising method of automatic word recognition in continuous speech, recently designated as 'word spotting', has been demonstrated. The method uses error residual ratios from LPC (Linear Predictive Coding) vocoder analysis for waveform comparison and a dynamic programming procedure for t...
Gespeichert in:
| 1. Verfasser: | |
|---|---|
| Format: | Abschlussarbeit Mikrofilm Buch |
| Sprache: | English |
| Veröffentlicht: |
1976
|
| Ausgabe: | [Mikrofiche-Ausg.] |
| Schlagworte: | |
| Zusammenfassung: | A promising method of automatic word recognition in continuous speech, recently designated as 'word spotting', has been demonstrated. The method uses error residual ratios from LPC (Linear Predictive Coding) vocoder analysis for waveform comparison and a dynamic programming procedure for time registration between the incoming speech and a template of the key word. Using a similarity threshold, the incoming speech is compared with several templates to account for variability in spectral shape. This system can work in real time using a real time vocoder. The multiple templates are used in such a way that a small number of templates, 3 or 4, is made to look like several hundred or more. This is accomplished by dynamically constructing a composite template from parts of each single template as part of the processing of the incoming speech so a particular composite template is constructed for each word being recognized. An accuracy of 99% with no false alarms was achieved using 205 key words, 5 different speakers, and approximately 10 minutes of speech text. Performance in the presence of additive white gaussian noise of approximately 11 dB signal-to-noise ratio was 66%; when the speech was processed to account for the noise, results improved to 85%-90% accuracy |
| Beschreibung: | IX, 78 S. graph. Darst. |
Internformat
MARC
| LEADER | 00000nam a2200000 c 4500 | ||
|---|---|---|---|
| 001 | BV035381003 | ||
| 003 | DE-604 | ||
| 005 | 00000000000000.0 | ||
| 007 | he|uuuuuuuucu | ||
| 008 | 090319s1976 d||| bm||| 00||| eng d | ||
| 035 | |a (OCoLC)227453345 | ||
| 035 | |a (DE-599)BVBBV035381003 | ||
| 040 | |a DE-604 |b ger |e rakwb | ||
| 041 | 0 | |a eng | |
| 049 | |a DE-706 | ||
| 100 | 1 | |a Christiansen, Richard Wesley |e Verfasser |4 aut | |
| 245 | 1 | 0 | |a Word recognition in continuous speech using linear prediction analysis |c Richard Wesley Christiansen |
| 250 | |a [Mikrofiche-Ausg.] | ||
| 264 | 1 | |c 1976 | |
| 300 | |a IX, 78 S. |b graph. Darst. | ||
| 336 | |b txt |2 rdacontent | ||
| 337 | |b h |2 rdamedia | ||
| 338 | |b he |2 rdacarrier | ||
| 502 | |a Salt Lake City, Utah, Univ., Diss., 1976 | ||
| 520 | |a A promising method of automatic word recognition in continuous speech, recently designated as 'word spotting', has been demonstrated. The method uses error residual ratios from LPC (Linear Predictive Coding) vocoder analysis for waveform comparison and a dynamic programming procedure for time registration between the incoming speech and a template of the key word. Using a similarity threshold, the incoming speech is compared with several templates to account for variability in spectral shape. This system can work in real time using a real time vocoder. The multiple templates are used in such a way that a small number of templates, 3 or 4, is made to look like several hundred or more. This is accomplished by dynamically constructing a composite template from parts of each single template as part of the processing of the incoming speech so a particular composite template is constructed for each word being recognized. An accuracy of 99% with no false alarms was achieved using 205 key words, 5 different speakers, and approximately 10 minutes of speech text. Performance in the presence of additive white gaussian noise of approximately 11 dB signal-to-noise ratio was 66%; when the speech was processed to account for the noise, results improved to 85%-90% accuracy | ||
| 533 | |a Mikroform-Ausgabe |b Ann Arbor, Mich. |c Univ. Microfilms Internat. |d 1977 |e 1 Mikrofiche |n Mikrofiche-Ausg.: |7 s1977 | ||
| 650 | 4 | |a Word spotting | |
| 650 | 4 | |a Linear predictive coding | |
| 650 | 4 | |a Error residual ratios | |
| 650 | 7 | |a Computer Programming and Software |2 scgdst | |
| 650 | 7 | |a Cybernetics |2 scgdst | |
| 650 | 7 | |a Voice Communications |2 scgdst | |
| 650 | 7 | |a Speech recognition |2 dtict | |
| 650 | 7 | |a Dynamic programming |2 dtict | |
| 650 | 7 | |a Algorithms |2 dtict | |
| 650 | 7 | |a Signal processing |2 dtict | |
| 650 | 7 | |a Threshold effects |2 dtict | |
| 650 | 7 | |a Comparison |2 dtict | |
| 650 | 7 | |a Waveforms |2 dtict | |
| 650 | 7 | |a Accuracy |2 dtict | |
| 650 | 7 | |a Fortran |2 dtict | |
| 650 | 7 | |a Mathematical prediction |2 dtict | |
| 650 | 7 | |a Error analysis |2 dtict | |
| 650 | 7 | |a Templates |2 dtict | |
| 650 | 7 | |a Words(language) |2 dtict | |
| 650 | 7 | |a Speech analysis |2 dtict | |
| 650 | 7 | |a Speech |2 dtict | |
| 650 | 7 | |a Spectrum analysis |2 dtict | |
| 650 | 7 | |a Vocoders |2 dtict | |
| 650 | 4 | |a Accuracy / dtict | |
| 650 | 4 | |a Algorithms / dtict | |
| 650 | 4 | |a Comparison / dtict | |
| 650 | 4 | |a Computer Programming and Software / scgdst | |
| 650 | 4 | |a Cybernetics / scgdst | |
| 650 | 4 | |a Dynamic programming / dtict | |
| 650 | 4 | |a Error analysis / dtict | |
| 650 | 4 | |a Error residual ratios | |
| 650 | 4 | |a Fortran / dtict | |
| 650 | 4 | |a Linear predictive coding | |
| 650 | 4 | |a Mathematical prediction / dtict | |
| 650 | 4 | |a Signal processing / dtict | |
| 650 | 4 | |a Spectrum analysis / dtict | |
| 650 | 4 | |a Speech / dtict | |
| 650 | 4 | |a Speech analysis / dtict | |
| 650 | 4 | |a Speech recognition / dtict | |
| 650 | 4 | |a Templates / dtict | |
| 650 | 4 | |a Threshold effects / dtict | |
| 650 | 4 | |a Vocoders / dtict | |
| 650 | 4 | |a Voice Communications / scgdst | |
| 650 | 4 | |a Waveforms / dtict | |
| 650 | 4 | |a Word spotting | |
| 650 | 4 | |a Words(language) / dtict | |
| 655 | 7 | |0 (DE-588)4113937-9 |a Hochschulschrift |2 gnd-content | |
| 776 | 0 | 8 | |i Reproduktion von |a Christiansen, Richard Wesley |t Word recognition in continuous speech using linear prediction analysis |d 1976 |
| 999 | |a oai:aleph.bib-bvb.de:BVB01-017185261 | ||
Datensatz im Suchindex
| _version_ | 1804138710507192320 |
|---|---|
| any_adam_object | |
| author | Christiansen, Richard Wesley |
| author_facet | Christiansen, Richard Wesley |
| author_role | aut |
| author_sort | Christiansen, Richard Wesley |
| author_variant | r w c rw rwc |
| building | Verbundindex |
| bvnumber | BV035381003 |
| ctrlnum | (OCoLC)227453345 (DE-599)BVBBV035381003 |
| edition | [Mikrofiche-Ausg.] |
| format | Thesis Microfilm Book |
| fullrecord | <?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>04204nam a2200865 c 4500</leader><controlfield tag="001">BV035381003</controlfield><controlfield tag="003">DE-604</controlfield><controlfield tag="005">00000000000000.0</controlfield><controlfield tag="007">he|uuuuuuuucu</controlfield><controlfield tag="008">090319s1976 d||| bm||| 00||| eng d</controlfield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(OCoLC)227453345</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-599)BVBBV035381003</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-604</subfield><subfield code="b">ger</subfield><subfield code="e">rakwb</subfield></datafield><datafield tag="041" ind1="0" ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="049" ind1=" " ind2=" "><subfield code="a">DE-706</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Christiansen, Richard Wesley</subfield><subfield code="e">Verfasser</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Word recognition in continuous speech using linear prediction analysis</subfield><subfield code="c">Richard Wesley Christiansen</subfield></datafield><datafield tag="250" ind1=" " ind2=" "><subfield code="a">[Mikrofiche-Ausg.]</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">1976</subfield></datafield><datafield tag="300" ind1=" " ind2=" "><subfield code="a">IX, 78 S.</subfield><subfield code="b">graph. Darst.</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="b">txt</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="b">h</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="b">he</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="502" ind1=" " ind2=" "><subfield code="a">Salt Lake City, Utah, Univ., Diss., 1976</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">A promising method of automatic word recognition in continuous speech, recently designated as 'word spotting', has been demonstrated. The method uses error residual ratios from LPC (Linear Predictive Coding) vocoder analysis for waveform comparison and a dynamic programming procedure for time registration between the incoming speech and a template of the key word. Using a similarity threshold, the incoming speech is compared with several templates to account for variability in spectral shape. This system can work in real time using a real time vocoder. The multiple templates are used in such a way that a small number of templates, 3 or 4, is made to look like several hundred or more. This is accomplished by dynamically constructing a composite template from parts of each single template as part of the processing of the incoming speech so a particular composite template is constructed for each word being recognized. An accuracy of 99% with no false alarms was achieved using 205 key words, 5 different speakers, and approximately 10 minutes of speech text. Performance in the presence of additive white gaussian noise of approximately 11 dB signal-to-noise ratio was 66%; when the speech was processed to account for the noise, results improved to 85%-90% accuracy</subfield></datafield><datafield tag="533" ind1=" " ind2=" "><subfield code="a">Mikroform-Ausgabe</subfield><subfield code="b">Ann Arbor, Mich.</subfield><subfield code="c">Univ. Microfilms Internat.</subfield><subfield code="d">1977</subfield><subfield code="e">1 Mikrofiche</subfield><subfield code="n">Mikrofiche-Ausg.:</subfield><subfield code="7">s1977</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Word spotting</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Linear predictive coding</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Error residual ratios</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Computer Programming and Software</subfield><subfield code="2">scgdst</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Cybernetics</subfield><subfield code="2">scgdst</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Voice Communications</subfield><subfield code="2">scgdst</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Speech recognition</subfield><subfield code="2">dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Dynamic programming</subfield><subfield code="2">dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Algorithms</subfield><subfield code="2">dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Signal processing</subfield><subfield code="2">dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Threshold effects</subfield><subfield code="2">dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Comparison</subfield><subfield code="2">dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Waveforms</subfield><subfield code="2">dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Accuracy</subfield><subfield code="2">dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Fortran</subfield><subfield code="2">dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Mathematical prediction</subfield><subfield code="2">dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Error analysis</subfield><subfield code="2">dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Templates</subfield><subfield code="2">dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Words(language)</subfield><subfield code="2">dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Speech analysis</subfield><subfield code="2">dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Speech</subfield><subfield code="2">dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Spectrum analysis</subfield><subfield code="2">dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Vocoders</subfield><subfield code="2">dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Accuracy / dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Algorithms / dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Comparison / dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Computer Programming and Software / scgdst</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Cybernetics / scgdst</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Dynamic programming / dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Error analysis / dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Error residual ratios</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Fortran / dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Linear predictive coding</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Mathematical prediction / dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Signal processing / dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Spectrum analysis / dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Speech / dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Speech analysis / dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Speech recognition / dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Templates / dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Threshold effects / dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Vocoders / dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Voice Communications / scgdst</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Waveforms / dtict</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Word spotting</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Words(language) / dtict</subfield></datafield><datafield tag="655" ind1=" " ind2="7"><subfield code="0">(DE-588)4113937-9</subfield><subfield code="a">Hochschulschrift</subfield><subfield code="2">gnd-content</subfield></datafield><datafield tag="776" ind1="0" ind2="8"><subfield code="i">Reproduktion von</subfield><subfield code="a">Christiansen, Richard Wesley</subfield><subfield code="t">Word recognition in continuous speech using linear prediction analysis</subfield><subfield code="d">1976</subfield></datafield><datafield tag="999" ind1=" " ind2=" "><subfield code="a">oai:aleph.bib-bvb.de:BVB01-017185261</subfield></datafield></record></collection> |
| genre | (DE-588)4113937-9 Hochschulschrift gnd-content |
| genre_facet | Hochschulschrift |
| id | DE-604.BV035381003 |
| illustrated | Illustrated |
| indexdate | 2024-07-09T21:32:31Z |
| institution | BVB |
| language | English |
| oai_aleph_id | oai:aleph.bib-bvb.de:BVB01-017185261 |
| oclc_num | 227453345 |
| open_access_boolean | |
| owner | DE-706 |
| owner_facet | DE-706 |
| physical | IX, 78 S. graph. Darst. |
| publishDate | 1976 |
| publishDateSearch | 1976 |
| publishDateSort | 1976 |
| record_format | marc |
| spelling | Christiansen, Richard Wesley Verfasser aut Word recognition in continuous speech using linear prediction analysis Richard Wesley Christiansen [Mikrofiche-Ausg.] 1976 IX, 78 S. graph. Darst. txt rdacontent h rdamedia he rdacarrier Salt Lake City, Utah, Univ., Diss., 1976 A promising method of automatic word recognition in continuous speech, recently designated as 'word spotting', has been demonstrated. The method uses error residual ratios from LPC (Linear Predictive Coding) vocoder analysis for waveform comparison and a dynamic programming procedure for time registration between the incoming speech and a template of the key word. Using a similarity threshold, the incoming speech is compared with several templates to account for variability in spectral shape. This system can work in real time using a real time vocoder. The multiple templates are used in such a way that a small number of templates, 3 or 4, is made to look like several hundred or more. This is accomplished by dynamically constructing a composite template from parts of each single template as part of the processing of the incoming speech so a particular composite template is constructed for each word being recognized. An accuracy of 99% with no false alarms was achieved using 205 key words, 5 different speakers, and approximately 10 minutes of speech text. Performance in the presence of additive white gaussian noise of approximately 11 dB signal-to-noise ratio was 66%; when the speech was processed to account for the noise, results improved to 85%-90% accuracy Mikroform-Ausgabe Ann Arbor, Mich. Univ. Microfilms Internat. 1977 1 Mikrofiche Mikrofiche-Ausg.: s1977 Word spotting Linear predictive coding Error residual ratios Computer Programming and Software scgdst Cybernetics scgdst Voice Communications scgdst Speech recognition dtict Dynamic programming dtict Algorithms dtict Signal processing dtict Threshold effects dtict Comparison dtict Waveforms dtict Accuracy dtict Fortran dtict Mathematical prediction dtict Error analysis dtict Templates dtict Words(language) dtict Speech analysis dtict Speech dtict Spectrum analysis dtict Vocoders dtict Accuracy / dtict Algorithms / dtict Comparison / dtict Computer Programming and Software / scgdst Cybernetics / scgdst Dynamic programming / dtict Error analysis / dtict Fortran / dtict Mathematical prediction / dtict Signal processing / dtict Spectrum analysis / dtict Speech / dtict Speech analysis / dtict Speech recognition / dtict Templates / dtict Threshold effects / dtict Vocoders / dtict Voice Communications / scgdst Waveforms / dtict Words(language) / dtict (DE-588)4113937-9 Hochschulschrift gnd-content Reproduktion von Christiansen, Richard Wesley Word recognition in continuous speech using linear prediction analysis 1976 |
| spellingShingle | Christiansen, Richard Wesley Word recognition in continuous speech using linear prediction analysis Word spotting Linear predictive coding Error residual ratios Computer Programming and Software scgdst Cybernetics scgdst Voice Communications scgdst Speech recognition dtict Dynamic programming dtict Algorithms dtict Signal processing dtict Threshold effects dtict Comparison dtict Waveforms dtict Accuracy dtict Fortran dtict Mathematical prediction dtict Error analysis dtict Templates dtict Words(language) dtict Speech analysis dtict Speech dtict Spectrum analysis dtict Vocoders dtict Accuracy / dtict Algorithms / dtict Comparison / dtict Computer Programming and Software / scgdst Cybernetics / scgdst Dynamic programming / dtict Error analysis / dtict Fortran / dtict Mathematical prediction / dtict Signal processing / dtict Spectrum analysis / dtict Speech / dtict Speech analysis / dtict Speech recognition / dtict Templates / dtict Threshold effects / dtict Vocoders / dtict Voice Communications / scgdst Waveforms / dtict Words(language) / dtict |
| subject_GND | (DE-588)4113937-9 |
| title | Word recognition in continuous speech using linear prediction analysis |
| title_auth | Word recognition in continuous speech using linear prediction analysis |
| title_exact_search | Word recognition in continuous speech using linear prediction analysis |
| title_full | Word recognition in continuous speech using linear prediction analysis Richard Wesley Christiansen |
| title_fullStr | Word recognition in continuous speech using linear prediction analysis Richard Wesley Christiansen |
| title_full_unstemmed | Word recognition in continuous speech using linear prediction analysis Richard Wesley Christiansen |
| title_short | Word recognition in continuous speech using linear prediction analysis |
| title_sort | word recognition in continuous speech using linear prediction analysis |
| topic | Word spotting Linear predictive coding Error residual ratios Computer Programming and Software scgdst Cybernetics scgdst Voice Communications scgdst Speech recognition dtict Dynamic programming dtict Algorithms dtict Signal processing dtict Threshold effects dtict Comparison dtict Waveforms dtict Accuracy dtict Fortran dtict Mathematical prediction dtict Error analysis dtict Templates dtict Words(language) dtict Speech analysis dtict Speech dtict Spectrum analysis dtict Vocoders dtict Accuracy / dtict Algorithms / dtict Comparison / dtict Computer Programming and Software / scgdst Cybernetics / scgdst Dynamic programming / dtict Error analysis / dtict Fortran / dtict Mathematical prediction / dtict Signal processing / dtict Spectrum analysis / dtict Speech / dtict Speech analysis / dtict Speech recognition / dtict Templates / dtict Threshold effects / dtict Vocoders / dtict Voice Communications / scgdst Waveforms / dtict Words(language) / dtict |
| topic_facet | Word spotting Linear predictive coding Error residual ratios Computer Programming and Software Cybernetics Voice Communications Speech recognition Dynamic programming Algorithms Signal processing Threshold effects Comparison Waveforms Accuracy Fortran Mathematical prediction Error analysis Templates Words(language) Speech analysis Speech Spectrum analysis Vocoders Accuracy / dtict Algorithms / dtict Comparison / dtict Computer Programming and Software / scgdst Cybernetics / scgdst Dynamic programming / dtict Error analysis / dtict Fortran / dtict Mathematical prediction / dtict Signal processing / dtict Spectrum analysis / dtict Speech / dtict Speech analysis / dtict Speech recognition / dtict Templates / dtict Threshold effects / dtict Vocoders / dtict Voice Communications / scgdst Waveforms / dtict Words(language) / dtict Hochschulschrift |
| work_keys_str_mv | AT christiansenrichardwesley wordrecognitionincontinuousspeechusinglinearpredictionanalysis |