Face Anti-Spoofing Method Based on Residual Network with Channel Attention Mechanism
The face recognition system is vulnerable to spoofing attacks by photos or videos of a valid user face. However, edge degradation and texture blurring occur when non-living face images are used to attack the face recognition system. With this in mind, a novel face anti-spoofing method combines the r...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-09-01
|
| Series: | Electronics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2079-9292/11/19/3056 |
| _version_ | 1797479851351867392 |
|---|---|
| author | Yueping Kong Xinyuan Li Guangye Hao Chu Liu |
| author_facet | Yueping Kong Xinyuan Li Guangye Hao Chu Liu |
| author_sort | Yueping Kong |
| collection | DOAJ |
| description | The face recognition system is vulnerable to spoofing attacks by photos or videos of a valid user face. However, edge degradation and texture blurring occur when non-living face images are used to attack the face recognition system. With this in mind, a novel face anti-spoofing method combines the residual network and the channel attention mechanism. In our method, the residual network extracts the texture differences of features between face images. In contrast, the attention mechanism focuses on the differences of shadow and edge features located on nasal and cheek areas between living and non-living face images. It can assign weights to different filter features of the face image and enhance the ability of network extraction and expression of different key features in the nasal and cheek regions, improving detection accuracy. The experiments were performed on the public face anti-spoofing datasets of Replay-Attack and CASIA-FASD. We found the best value of the parameter <i>r</i> suitable for face anti-spoofing research is 16, and the accuracy of the method is 99.98% and 97.75%, respectively. Furthermore, to enhance the robustness of the method to illumination changes, the experiment was also performed on the datasets with light changes and achieved a good result. |
| first_indexed | 2024-03-09T21:51:43Z |
| format | Article |
| id | doaj.art-4e7601224e6c492eb916f9eb0e3e54d1 |
| institution | Directory Open Access Journal |
| issn | 2079-9292 |
| language | English |
| last_indexed | 2024-03-09T21:51:43Z |
| publishDate | 2022-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Electronics |
| spelling | doaj.art-4e7601224e6c492eb916f9eb0e3e54d12023-11-23T20:05:34ZengMDPI AGElectronics2079-92922022-09-011119305610.3390/electronics11193056Face Anti-Spoofing Method Based on Residual Network with Channel Attention MechanismYueping Kong0Xinyuan Li1Guangye Hao2Chu Liu3School of Information and Control Engineering, Xi’an University of Architecture and Technology, Xi’an 710055, ChinaSchool of Information and Control Engineering, Xi’an University of Architecture and Technology, Xi’an 710055, ChinaInformatization Technology Office, Shaanxi Provincial Public Security Department, Xi’an 710018, ChinaSchool of Information and Control Engineering, Xi’an University of Architecture and Technology, Xi’an 710055, ChinaThe face recognition system is vulnerable to spoofing attacks by photos or videos of a valid user face. However, edge degradation and texture blurring occur when non-living face images are used to attack the face recognition system. With this in mind, a novel face anti-spoofing method combines the residual network and the channel attention mechanism. In our method, the residual network extracts the texture differences of features between face images. In contrast, the attention mechanism focuses on the differences of shadow and edge features located on nasal and cheek areas between living and non-living face images. It can assign weights to different filter features of the face image and enhance the ability of network extraction and expression of different key features in the nasal and cheek regions, improving detection accuracy. The experiments were performed on the public face anti-spoofing datasets of Replay-Attack and CASIA-FASD. We found the best value of the parameter <i>r</i> suitable for face anti-spoofing research is 16, and the accuracy of the method is 99.98% and 97.75%, respectively. Furthermore, to enhance the robustness of the method to illumination changes, the experiment was also performed on the datasets with light changes and achieved a good result.https://www.mdpi.com/2079-9292/11/19/3056face anti-spoofingsecondary imagingresidual networkattention mechanism |
| spellingShingle | Yueping Kong Xinyuan Li Guangye Hao Chu Liu Face Anti-Spoofing Method Based on Residual Network with Channel Attention Mechanism Electronics face anti-spoofing secondary imaging residual network attention mechanism |
| title | Face Anti-Spoofing Method Based on Residual Network with Channel Attention Mechanism |
| title_full | Face Anti-Spoofing Method Based on Residual Network with Channel Attention Mechanism |
| title_fullStr | Face Anti-Spoofing Method Based on Residual Network with Channel Attention Mechanism |
| title_full_unstemmed | Face Anti-Spoofing Method Based on Residual Network with Channel Attention Mechanism |
| title_short | Face Anti-Spoofing Method Based on Residual Network with Channel Attention Mechanism |
| title_sort | face anti spoofing method based on residual network with channel attention mechanism |
| topic | face anti-spoofing secondary imaging residual network attention mechanism |
| url | https://www.mdpi.com/2079-9292/11/19/3056 |
| work_keys_str_mv | AT yuepingkong faceantispoofingmethodbasedonresidualnetworkwithchannelattentionmechanism AT xinyuanli faceantispoofingmethodbasedonresidualnetworkwithchannelattentionmechanism AT guangyehao faceantispoofingmethodbasedonresidualnetworkwithchannelattentionmechanism AT chuliu faceantispoofingmethodbasedonresidualnetworkwithchannelattentionmechanism |