Spoofing detection from a feature representation perspective
Spoofing detection, which discriminates the spoofed speech from the natural speech, has gained much attention recently. Low-dimensional features that are used in speaker recognition/verification are also used in spoofing detection. Unfortunately, they don't capture sufficient information requir...
| Main Authors: | , , , , |
|---|---|
| Other Authors: | |
| Format: | Conference Paper |
| Language: | English |
| Published: |
2018
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/89643 http://hdl.handle.net/10220/47063 |
| _version_ | 1824456013785333760 |
|---|---|
| author | Tian, Xiaohai Wu, Zhizheg Xiao, Xiong Chng, Eng Siong Li, Haizhou |
| author2 | School of Computer Science and Engineering |
| author_facet | School of Computer Science and Engineering Tian, Xiaohai Wu, Zhizheg Xiao, Xiong Chng, Eng Siong Li, Haizhou |
| author_sort | Tian, Xiaohai |
| collection | NTU |
| description | Spoofing detection, which discriminates the spoofed speech from the natural speech, has gained much attention recently. Low-dimensional features that are used in speaker recognition/verification are also used in spoofing detection. Unfortunately, they don't capture sufficient information required for spoofing detection. In this work, we investigate the use of high-dimensional features for spoofing detection, that maybe more sensitive to the artifacts in the spoofed speech. Six types of high-dimensional feature are employed. For each kind of feature, four different representations are extracted, i.e. the original high-dimensional feature, corresponding low-dimensional feature, the low- and the high-frequency regions of the original high-dimensional feature. Dynamic features are also calculated to assess the effectiveness of the temporal information to detect the artifacts across frames. A neural network-based classifier is adopted to handle the high-dimensional features. Experimental results on the standard ASVspoof 2015 corpus suggest that high-dimensional features and dynamic features are useful for spoofing attack detection. A fusion of them has been shown to achieve 0.0% the equal error rates for nine of ten attack types. |
| first_indexed | 2025-02-19T03:47:21Z |
| format | Conference Paper |
| id | ntu-10356/89643 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2025-02-19T03:47:21Z |
| publishDate | 2018 |
| record_format | dspace |
| spelling | ntu-10356/896432020-03-07T11:48:46Z Spoofing detection from a feature representation perspective Tian, Xiaohai Wu, Zhizheg Xiao, Xiong Chng, Eng Siong Li, Haizhou School of Computer Science and Engineering 2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) NTU-UBC Research Centre of Excellence in Active Living for the Elderly Temasek Laboratories Spoofing Detection DRNTU::Engineering::Computer science and engineering Spoofing Attack Spoofing detection, which discriminates the spoofed speech from the natural speech, has gained much attention recently. Low-dimensional features that are used in speaker recognition/verification are also used in spoofing detection. Unfortunately, they don't capture sufficient information required for spoofing detection. In this work, we investigate the use of high-dimensional features for spoofing detection, that maybe more sensitive to the artifacts in the spoofed speech. Six types of high-dimensional feature are employed. For each kind of feature, four different representations are extracted, i.e. the original high-dimensional feature, corresponding low-dimensional feature, the low- and the high-frequency regions of the original high-dimensional feature. Dynamic features are also calculated to assess the effectiveness of the temporal information to detect the artifacts across frames. A neural network-based classifier is adopted to handle the high-dimensional features. Experimental results on the standard ASVspoof 2015 corpus suggest that high-dimensional features and dynamic features are useful for spoofing attack detection. A fusion of them has been shown to achieve 0.0% the equal error rates for nine of ten attack types. NRF (Natl Research Foundation, S’pore) Accepted version 2018-12-18T07:34:06Z 2019-12-06T17:30:08Z 2018-12-18T07:34:06Z 2019-12-06T17:30:08Z 2016-03-01 2016 Conference Paper Tian, X., Wu, Z., Xiao, X., Chng, E. S., & Li, H. (2016). Spoofing detection from a feature representation perspective. 2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2119-2123. doi:10.1109/ICASSP.2016.7472051 https://hdl.handle.net/10356/89643 http://hdl.handle.net/10220/47063 10.1109/ICASSP.2016.7472051 200443 en © 2016 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. The published version is available at: [http://dx.doi.org/10.1109/ICASSP.2016.7472051]. 5 p. application/pdf |
| spellingShingle | Spoofing Detection DRNTU::Engineering::Computer science and engineering Spoofing Attack Tian, Xiaohai Wu, Zhizheg Xiao, Xiong Chng, Eng Siong Li, Haizhou Spoofing detection from a feature representation perspective |
| title | Spoofing detection from a feature representation perspective |
| title_full | Spoofing detection from a feature representation perspective |
| title_fullStr | Spoofing detection from a feature representation perspective |
| title_full_unstemmed | Spoofing detection from a feature representation perspective |
| title_short | Spoofing detection from a feature representation perspective |
| title_sort | spoofing detection from a feature representation perspective |
| topic | Spoofing Detection DRNTU::Engineering::Computer science and engineering Spoofing Attack |
| url | https://hdl.handle.net/10356/89643 http://hdl.handle.net/10220/47063 |
| work_keys_str_mv | AT tianxiaohai spoofingdetectionfromafeaturerepresentationperspective AT wuzhizheg spoofingdetectionfromafeaturerepresentationperspective AT xiaoxiong spoofingdetectionfromafeaturerepresentationperspective AT chngengsiong spoofingdetectionfromafeaturerepresentationperspective AT lihaizhou spoofingdetectionfromafeaturerepresentationperspective |