Uncertainty assessment for detection of spoofing attacks to speaker verification systems using a Bayesian approach
There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the m...
Ausführliche Beschreibung
Gespeichert in:
| Autor*in: |
Süslü, Çağıl [verfasserIn] |
|---|
| Format: |
E-Artikel |
|---|---|
| Sprache: |
Englisch |
| Erschienen: |
2022transfer abstract |
|---|
| Schlagwörter: |
|---|
| Umfang: |
8 |
|---|
| Übergeordnetes Werk: |
Enthalten in: Comparison of dosing algorithms for acenocoumarol and phenprocoumon using clinical factors with the standard care in the Netherlands - Zhang, Yumao ELSEVIER, 2015, an interdisciplinary journal, Amsterdam |
|---|---|
| Übergeordnetes Werk: |
volume:137 ; year:2022 ; pages:44-51 ; extent:8 |
| Links: |
|---|
| DOI / URN: |
10.1016/j.specom.2021.12.003 |
|---|
| Katalog-ID: |
ELV056752423 |
|---|
| LEADER | 01000caa a22002652 4500 | ||
|---|---|---|---|
| 001 | ELV056752423 | ||
| 003 | DE-627 | ||
| 005 | 20230626043912.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 220808s2022 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.1016/j.specom.2021.12.003 |2 doi | |
| 028 | 5 | 2 | |a /cbs_pica/cbs_olc/import_discovery/elsevier/einzuspielen/GBV00000000001670.pica |
| 035 | |a (DE-627)ELV056752423 | ||
| 035 | |a (ELSEVIER)S0167-6393(21)00136-9 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 082 | 0 | 4 | |a 610 |q VZ |
| 082 | 0 | 4 | |a 610 |q VZ |
| 082 | 0 | 4 | |a 610 |q VZ |
| 084 | |a 44.40 |2 bkl | ||
| 100 | 1 | |a Süslü, Çağıl |e verfasserin |4 aut | |
| 245 | 1 | 0 | |a Uncertainty assessment for detection of spoofing attacks to speaker verification systems using a Bayesian approach |
| 264 | 1 | |c 2022transfer abstract | |
| 300 | |a 8 | ||
| 336 | |a nicht spezifiziert |b zzz |2 rdacontent | ||
| 337 | |a nicht spezifiziert |b z |2 rdamedia | ||
| 338 | |a nicht spezifiziert |b zu |2 rdacarrier | ||
| 520 | |a There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold. | ||
| 520 | |a There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold. | ||
| 650 | 7 | |a Uncertainty |2 Elsevier | |
| 650 | 7 | |a Spoofing countermeasure system |2 Elsevier | |
| 650 | 7 | |a Bayesian |2 Elsevier | |
| 650 | 7 | |a Speech |2 Elsevier | |
| 650 | 7 | |a Automatic speaker verification |2 Elsevier | |
| 700 | 1 | |a Eren, Eray |4 oth | |
| 700 | 1 | |a Demiroğlu, Cenk |4 oth | |
| 773 | 0 | 8 | |i Enthalten in |n North-Holland Publ. Comp |a Zhang, Yumao ELSEVIER |t Comparison of dosing algorithms for acenocoumarol and phenprocoumon using clinical factors with the standard care in the Netherlands |d 2015 |d an interdisciplinary journal |g Amsterdam |w (DE-627)ELV024100463 |
| 773 | 1 | 8 | |g volume:137 |g year:2022 |g pages:44-51 |g extent:8 |
| 856 | 4 | 0 | |u https://doi.org/10.1016/j.specom.2021.12.003 |3 Volltext |
| 912 | |a GBV_USEFLAG_U | ||
| 912 | |a GBV_ELV | ||
| 912 | |a SYSFLAG_U | ||
| 912 | |a SSG-OLC-PHA | ||
| 912 | |a SSG-OPC-PHA | ||
| 912 | |a GBV_ILN_40 | ||
| 936 | b | k | |a 44.40 |j Pharmazie |j Pharmazeutika |q VZ |
| 951 | |a AR | ||
| 952 | |d 137 |j 2022 |h 44-51 |g 8 | ||
| author_variant |
ç s çs |
|---|---|
| matchkey_str |
sslalereneraydemirolucenk:2022----:netitassmnfreetoosofnatcsopaevrfctoss |
| hierarchy_sort_str |
2022transfer abstract |
| bklnumber |
44.40 |
| publishDate |
2022 |
| allfields |
10.1016/j.specom.2021.12.003 doi /cbs_pica/cbs_olc/import_discovery/elsevier/einzuspielen/GBV00000000001670.pica (DE-627)ELV056752423 (ELSEVIER)S0167-6393(21)00136-9 DE-627 ger DE-627 rakwb eng 610 VZ 610 VZ 610 VZ 44.40 bkl Süslü, Çağıl verfasserin aut Uncertainty assessment for detection of spoofing attacks to speaker verification systems using a Bayesian approach 2022transfer abstract 8 nicht spezifiziert zzz rdacontent nicht spezifiziert z rdamedia nicht spezifiziert zu rdacarrier There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold. There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold. Uncertainty Elsevier Spoofing countermeasure system Elsevier Bayesian Elsevier Speech Elsevier Automatic speaker verification Elsevier Eren, Eray oth Demiroğlu, Cenk oth Enthalten in North-Holland Publ. Comp Zhang, Yumao ELSEVIER Comparison of dosing algorithms for acenocoumarol and phenprocoumon using clinical factors with the standard care in the Netherlands 2015 an interdisciplinary journal Amsterdam (DE-627)ELV024100463 volume:137 year:2022 pages:44-51 extent:8 https://doi.org/10.1016/j.specom.2021.12.003 Volltext GBV_USEFLAG_U GBV_ELV SYSFLAG_U SSG-OLC-PHA SSG-OPC-PHA GBV_ILN_40 44.40 Pharmazie Pharmazeutika VZ AR 137 2022 44-51 8 |
| spelling |
10.1016/j.specom.2021.12.003 doi /cbs_pica/cbs_olc/import_discovery/elsevier/einzuspielen/GBV00000000001670.pica (DE-627)ELV056752423 (ELSEVIER)S0167-6393(21)00136-9 DE-627 ger DE-627 rakwb eng 610 VZ 610 VZ 610 VZ 44.40 bkl Süslü, Çağıl verfasserin aut Uncertainty assessment for detection of spoofing attacks to speaker verification systems using a Bayesian approach 2022transfer abstract 8 nicht spezifiziert zzz rdacontent nicht spezifiziert z rdamedia nicht spezifiziert zu rdacarrier There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold. There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold. Uncertainty Elsevier Spoofing countermeasure system Elsevier Bayesian Elsevier Speech Elsevier Automatic speaker verification Elsevier Eren, Eray oth Demiroğlu, Cenk oth Enthalten in North-Holland Publ. Comp Zhang, Yumao ELSEVIER Comparison of dosing algorithms for acenocoumarol and phenprocoumon using clinical factors with the standard care in the Netherlands 2015 an interdisciplinary journal Amsterdam (DE-627)ELV024100463 volume:137 year:2022 pages:44-51 extent:8 https://doi.org/10.1016/j.specom.2021.12.003 Volltext GBV_USEFLAG_U GBV_ELV SYSFLAG_U SSG-OLC-PHA SSG-OPC-PHA GBV_ILN_40 44.40 Pharmazie Pharmazeutika VZ AR 137 2022 44-51 8 |
| allfields_unstemmed |
10.1016/j.specom.2021.12.003 doi /cbs_pica/cbs_olc/import_discovery/elsevier/einzuspielen/GBV00000000001670.pica (DE-627)ELV056752423 (ELSEVIER)S0167-6393(21)00136-9 DE-627 ger DE-627 rakwb eng 610 VZ 610 VZ 610 VZ 44.40 bkl Süslü, Çağıl verfasserin aut Uncertainty assessment for detection of spoofing attacks to speaker verification systems using a Bayesian approach 2022transfer abstract 8 nicht spezifiziert zzz rdacontent nicht spezifiziert z rdamedia nicht spezifiziert zu rdacarrier There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold. There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold. Uncertainty Elsevier Spoofing countermeasure system Elsevier Bayesian Elsevier Speech Elsevier Automatic speaker verification Elsevier Eren, Eray oth Demiroğlu, Cenk oth Enthalten in North-Holland Publ. Comp Zhang, Yumao ELSEVIER Comparison of dosing algorithms for acenocoumarol and phenprocoumon using clinical factors with the standard care in the Netherlands 2015 an interdisciplinary journal Amsterdam (DE-627)ELV024100463 volume:137 year:2022 pages:44-51 extent:8 https://doi.org/10.1016/j.specom.2021.12.003 Volltext GBV_USEFLAG_U GBV_ELV SYSFLAG_U SSG-OLC-PHA SSG-OPC-PHA GBV_ILN_40 44.40 Pharmazie Pharmazeutika VZ AR 137 2022 44-51 8 |
| allfieldsGer |
10.1016/j.specom.2021.12.003 doi /cbs_pica/cbs_olc/import_discovery/elsevier/einzuspielen/GBV00000000001670.pica (DE-627)ELV056752423 (ELSEVIER)S0167-6393(21)00136-9 DE-627 ger DE-627 rakwb eng 610 VZ 610 VZ 610 VZ 44.40 bkl Süslü, Çağıl verfasserin aut Uncertainty assessment for detection of spoofing attacks to speaker verification systems using a Bayesian approach 2022transfer abstract 8 nicht spezifiziert zzz rdacontent nicht spezifiziert z rdamedia nicht spezifiziert zu rdacarrier There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold. There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold. Uncertainty Elsevier Spoofing countermeasure system Elsevier Bayesian Elsevier Speech Elsevier Automatic speaker verification Elsevier Eren, Eray oth Demiroğlu, Cenk oth Enthalten in North-Holland Publ. Comp Zhang, Yumao ELSEVIER Comparison of dosing algorithms for acenocoumarol and phenprocoumon using clinical factors with the standard care in the Netherlands 2015 an interdisciplinary journal Amsterdam (DE-627)ELV024100463 volume:137 year:2022 pages:44-51 extent:8 https://doi.org/10.1016/j.specom.2021.12.003 Volltext GBV_USEFLAG_U GBV_ELV SYSFLAG_U SSG-OLC-PHA SSG-OPC-PHA GBV_ILN_40 44.40 Pharmazie Pharmazeutika VZ AR 137 2022 44-51 8 |
| allfieldsSound |
10.1016/j.specom.2021.12.003 doi /cbs_pica/cbs_olc/import_discovery/elsevier/einzuspielen/GBV00000000001670.pica (DE-627)ELV056752423 (ELSEVIER)S0167-6393(21)00136-9 DE-627 ger DE-627 rakwb eng 610 VZ 610 VZ 610 VZ 44.40 bkl Süslü, Çağıl verfasserin aut Uncertainty assessment for detection of spoofing attacks to speaker verification systems using a Bayesian approach 2022transfer abstract 8 nicht spezifiziert zzz rdacontent nicht spezifiziert z rdamedia nicht spezifiziert zu rdacarrier There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold. There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold. Uncertainty Elsevier Spoofing countermeasure system Elsevier Bayesian Elsevier Speech Elsevier Automatic speaker verification Elsevier Eren, Eray oth Demiroğlu, Cenk oth Enthalten in North-Holland Publ. Comp Zhang, Yumao ELSEVIER Comparison of dosing algorithms for acenocoumarol and phenprocoumon using clinical factors with the standard care in the Netherlands 2015 an interdisciplinary journal Amsterdam (DE-627)ELV024100463 volume:137 year:2022 pages:44-51 extent:8 https://doi.org/10.1016/j.specom.2021.12.003 Volltext GBV_USEFLAG_U GBV_ELV SYSFLAG_U SSG-OLC-PHA SSG-OPC-PHA GBV_ILN_40 44.40 Pharmazie Pharmazeutika VZ AR 137 2022 44-51 8 |
| language |
English |
| source |
Enthalten in Comparison of dosing algorithms for acenocoumarol and phenprocoumon using clinical factors with the standard care in the Netherlands Amsterdam volume:137 year:2022 pages:44-51 extent:8 |
| sourceStr |
Enthalten in Comparison of dosing algorithms for acenocoumarol and phenprocoumon using clinical factors with the standard care in the Netherlands Amsterdam volume:137 year:2022 pages:44-51 extent:8 |
| format_phy_str_mv |
Article |
| bklname |
Pharmazie Pharmazeutika |
| institution |
findex.gbv.de |
| topic_facet |
Uncertainty Spoofing countermeasure system Bayesian Speech Automatic speaker verification |
| dewey-raw |
610 |
| isfreeaccess_bool |
false |
| container_title |
Comparison of dosing algorithms for acenocoumarol and phenprocoumon using clinical factors with the standard care in the Netherlands |
| authorswithroles_txt_mv |
Süslü, Çağıl @@aut@@ Eren, Eray @@oth@@ Demiroğlu, Cenk @@oth@@ |
| publishDateDaySort_date |
2022-01-01T00:00:00Z |
| hierarchy_top_id |
ELV024100463 |
| dewey-sort |
3610 |
| id |
ELV056752423 |
| language_de |
englisch |
| fullrecord |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000caa a22002652 4500</leader><controlfield tag="001">ELV056752423</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20230626043912.0</controlfield><controlfield tag="007">cr uuu---uuuuu</controlfield><controlfield tag="008">220808s2022 xx |||||o 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1016/j.specom.2021.12.003</subfield><subfield code="2">doi</subfield></datafield><datafield tag="028" ind1="5" ind2="2"><subfield code="a">/cbs_pica/cbs_olc/import_discovery/elsevier/einzuspielen/GBV00000000001670.pica</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)ELV056752423</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(ELSEVIER)S0167-6393(21)00136-9</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rakwb</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">610</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">610</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">610</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">44.40</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Süslü, Çağıl</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Uncertainty assessment for detection of spoofing attacks to speaker verification systems using a Bayesian approach</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2022transfer abstract</subfield></datafield><datafield tag="300" ind1=" " ind2=" "><subfield code="a">8</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">nicht spezifiziert</subfield><subfield code="b">zzz</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">nicht spezifiziert</subfield><subfield code="b">z</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">nicht spezifiziert</subfield><subfield code="b">zu</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold.</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold.</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Uncertainty</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Spoofing countermeasure system</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Bayesian</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Speech</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Automatic speaker verification</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Eren, Eray</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Demiroğlu, Cenk</subfield><subfield code="4">oth</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="n">North-Holland Publ. Comp</subfield><subfield code="a">Zhang, Yumao ELSEVIER</subfield><subfield code="t">Comparison of dosing algorithms for acenocoumarol and phenprocoumon using clinical factors with the standard care in the Netherlands</subfield><subfield code="d">2015</subfield><subfield code="d">an interdisciplinary journal</subfield><subfield code="g">Amsterdam</subfield><subfield code="w">(DE-627)ELV024100463</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:137</subfield><subfield code="g">year:2022</subfield><subfield code="g">pages:44-51</subfield><subfield code="g">extent:8</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://doi.org/10.1016/j.specom.2021.12.003</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_USEFLAG_U</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ELV</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_U</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-PHA</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OPC-PHA</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_40</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">44.40</subfield><subfield code="j">Pharmazie</subfield><subfield code="j">Pharmazeutika</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">137</subfield><subfield code="j">2022</subfield><subfield code="h">44-51</subfield><subfield code="g">8</subfield></datafield></record></collection>
|
| author |
Süslü, Çağıl |
| spellingShingle |
Süslü, Çağıl ddc 610 bkl 44.40 Elsevier Uncertainty Elsevier Spoofing countermeasure system Elsevier Bayesian Elsevier Speech Elsevier Automatic speaker verification Uncertainty assessment for detection of spoofing attacks to speaker verification systems using a Bayesian approach |
| authorStr |
Süslü, Çağıl |
| ppnlink_with_tag_str_mv |
@@773@@(DE-627)ELV024100463 |
| format |
electronic Article |
| dewey-ones |
610 - Medicine & health |
| delete_txt_mv |
keep |
| author_role |
aut |
| collection |
elsevier |
| remote_str |
true |
| illustrated |
Not Illustrated |
| topic_title |
610 VZ 44.40 bkl Uncertainty assessment for detection of spoofing attacks to speaker verification systems using a Bayesian approach Uncertainty Elsevier Spoofing countermeasure system Elsevier Bayesian Elsevier Speech Elsevier Automatic speaker verification Elsevier |
| topic |
ddc 610 bkl 44.40 Elsevier Uncertainty Elsevier Spoofing countermeasure system Elsevier Bayesian Elsevier Speech Elsevier Automatic speaker verification |
| topic_unstemmed |
ddc 610 bkl 44.40 Elsevier Uncertainty Elsevier Spoofing countermeasure system Elsevier Bayesian Elsevier Speech Elsevier Automatic speaker verification |
| topic_browse |
ddc 610 bkl 44.40 Elsevier Uncertainty Elsevier Spoofing countermeasure system Elsevier Bayesian Elsevier Speech Elsevier Automatic speaker verification |
| format_facet |
Elektronische Aufsätze Aufsätze Elektronische Ressource |
| format_main_str_mv |
Text Zeitschrift/Artikel |
| carriertype_str_mv |
zu |
| author2_variant |
e e ee c d cd |
| hierarchy_parent_title |
Comparison of dosing algorithms for acenocoumarol and phenprocoumon using clinical factors with the standard care in the Netherlands |
| hierarchy_parent_id |
ELV024100463 |
| dewey-tens |
610 - Medicine & health |
| hierarchy_top_title |
Comparison of dosing algorithms for acenocoumarol and phenprocoumon using clinical factors with the standard care in the Netherlands |
| isfreeaccess_txt |
false |
| familylinks_str_mv |
(DE-627)ELV024100463 |
| title |
Uncertainty assessment for detection of spoofing attacks to speaker verification systems using a Bayesian approach |
| ctrlnum |
(DE-627)ELV056752423 (ELSEVIER)S0167-6393(21)00136-9 |
| title_full |
Uncertainty assessment for detection of spoofing attacks to speaker verification systems using a Bayesian approach |
| author_sort |
Süslü, Çağıl |
| journal |
Comparison of dosing algorithms for acenocoumarol and phenprocoumon using clinical factors with the standard care in the Netherlands |
| journalStr |
Comparison of dosing algorithms for acenocoumarol and phenprocoumon using clinical factors with the standard care in the Netherlands |
| lang_code |
eng |
| isOA_bool |
false |
| dewey-hundreds |
600 - Technology |
| recordtype |
marc |
| publishDateSort |
2022 |
| contenttype_str_mv |
zzz |
| container_start_page |
44 |
| author_browse |
Süslü, Çağıl |
| container_volume |
137 |
| physical |
8 |
| class |
610 VZ 44.40 bkl |
| format_se |
Elektronische Aufsätze |
| author-letter |
Süslü, Çağıl |
| doi_str_mv |
10.1016/j.specom.2021.12.003 |
| dewey-full |
610 |
| title_sort |
uncertainty assessment for detection of spoofing attacks to speaker verification systems using a bayesian approach |
| title_auth |
Uncertainty assessment for detection of spoofing attacks to speaker verification systems using a Bayesian approach |
| abstract |
There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold. |
| abstractGer |
There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold. |
| abstract_unstemmed |
There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold. |
| collection_details |
GBV_USEFLAG_U GBV_ELV SYSFLAG_U SSG-OLC-PHA SSG-OPC-PHA GBV_ILN_40 |
| title_short |
Uncertainty assessment for detection of spoofing attacks to speaker verification systems using a Bayesian approach |
| url |
https://doi.org/10.1016/j.specom.2021.12.003 |
| remote_bool |
true |
| author2 |
Eren, Eray Demiroğlu, Cenk |
| author2Str |
Eren, Eray Demiroğlu, Cenk |
| ppnlink |
ELV024100463 |
| mediatype_str_mv |
z |
| isOA_txt |
false |
| hochschulschrift_bool |
false |
| author2_role |
oth oth |
| doi_str |
10.1016/j.specom.2021.12.003 |
| up_date |
2024-07-06T21:18:03.597Z |
| _version_ |
1803866009604456448 |
| fullrecord_marcxml |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000caa a22002652 4500</leader><controlfield tag="001">ELV056752423</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20230626043912.0</controlfield><controlfield tag="007">cr uuu---uuuuu</controlfield><controlfield tag="008">220808s2022 xx |||||o 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1016/j.specom.2021.12.003</subfield><subfield code="2">doi</subfield></datafield><datafield tag="028" ind1="5" ind2="2"><subfield code="a">/cbs_pica/cbs_olc/import_discovery/elsevier/einzuspielen/GBV00000000001670.pica</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)ELV056752423</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(ELSEVIER)S0167-6393(21)00136-9</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rakwb</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">610</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">610</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">610</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">44.40</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Süslü, Çağıl</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Uncertainty assessment for detection of spoofing attacks to speaker verification systems using a Bayesian approach</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2022transfer abstract</subfield></datafield><datafield tag="300" ind1=" " ind2=" "><subfield code="a">8</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">nicht spezifiziert</subfield><subfield code="b">zzz</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">nicht spezifiziert</subfield><subfield code="b">z</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">nicht spezifiziert</subfield><subfield code="b">zu</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold.</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">There has been tremendous progress in automatic speaker verification systems over the last decade. Still, spoofing attacks pose a significant challenge to their deployment. Even though there are various attack techniques such as voice conversion and speech synthesis, replay attacks pose one of the most important types since they can be done without significant expertise in speech technology. Moreover, replay attacks are hard to detect because they are done with simple replay of the original audio. The problem has gained more attention since the introduction of the ASV spoof 2017 challenge, which included a well-designed database with realistic replay attack conditions. Even though many different deep network types and acoustic features were proposed since the challenge, one key issue, which is model uncertainty around the neural networks’ decision is largely ignored. This is a result of using the softmax function with the cross-entropy loss, which is widely used in many domains. Here, we propose using evidential deep learning, which is a recently proposed method that is rapidly gaining popularity, for assessing the model uncertainty around the network’s decision. Experimental results show that the investigated network architectures perform better in terms of equal error rate with the new loss function. Moreover, reliability of measured uncertainty is shown by filtering samples out of the test set using the Bayesian uncertainty measure, which resulted with a consistent decrease in EER with decreasing threshold.</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Uncertainty</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Spoofing countermeasure system</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Bayesian</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Speech</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Automatic speaker verification</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Eren, Eray</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Demiroğlu, Cenk</subfield><subfield code="4">oth</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="n">North-Holland Publ. Comp</subfield><subfield code="a">Zhang, Yumao ELSEVIER</subfield><subfield code="t">Comparison of dosing algorithms for acenocoumarol and phenprocoumon using clinical factors with the standard care in the Netherlands</subfield><subfield code="d">2015</subfield><subfield code="d">an interdisciplinary journal</subfield><subfield code="g">Amsterdam</subfield><subfield code="w">(DE-627)ELV024100463</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:137</subfield><subfield code="g">year:2022</subfield><subfield code="g">pages:44-51</subfield><subfield code="g">extent:8</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://doi.org/10.1016/j.specom.2021.12.003</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_USEFLAG_U</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ELV</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_U</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-PHA</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OPC-PHA</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_40</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">44.40</subfield><subfield code="j">Pharmazie</subfield><subfield code="j">Pharmazeutika</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">137</subfield><subfield code="j">2022</subfield><subfield code="h">44-51</subfield><subfield code="g">8</subfield></datafield></record></collection>
|
| score |
7.4013834 |