CloudNet: A LiDAR-Based Face Anti-Spoofing Model That Is Robust Against Light Variation
Face anti-spoofing (FAS) is a technology that protects face recognition systems from presentation attacks. The current challenge faced by FAS studies is the difficulty in creating a generalized light variation model. This is because face data are sensitive to light domain. FAS models using only red...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10038600/ |
| _version_ | 1797894975357190144 |
|---|---|
| author | Yongrae Kim Hyunmin Gwak Jaehoon Oh Minho Kang Jinkyu Kim Hyun Kwon Sunghwan Kim |
| author_facet | Yongrae Kim Hyunmin Gwak Jaehoon Oh Minho Kang Jinkyu Kim Hyun Kwon Sunghwan Kim |
| author_sort | Yongrae Kim |
| collection | DOAJ |
| description | Face anti-spoofing (FAS) is a technology that protects face recognition systems from presentation attacks. The current challenge faced by FAS studies is the difficulty in creating a generalized light variation model. This is because face data are sensitive to light domain. FAS models using only red green blue (RGB) images suffer from poor performance when the training and test datasets have different light variations. To overcome this problem, this study focuses on light detection and ranging (LiDAR) sensors. LiDAR is a time-of-flight depth sensor that is included in the latest mobile devices. It is negligibly affected by light and provides 3D coordinate and depth information of the target. Thus, a model that is resistant to light variations and exhibiting excellent performance can be created. For the experiment, datasets collected with a LiDAR camera are built and CloudNet architectures for RGB, point clouds, and depth are designed. Three protocols are used to confirm the performance of the model according to variations in the light domain. Experimental results indicate that for protocols 2 and 3, CloudNet error rates increase by 0.1340 and 0.1528, whereas the error rates of the RGB model increase by 0.3951 and 0.4111, respectively, as compared with protocol 1. These results demonstrate that the LiDAR-based FAS model with CloudNet has a more generalized performance compared with the RGB model. |
| first_indexed | 2024-04-10T07:19:06Z |
| format | Article |
| id | doaj.art-f598ba821cd74de5a37851554f4250af |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-04-10T07:19:06Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-f598ba821cd74de5a37851554f4250af2023-02-25T00:02:23ZengIEEEIEEE Access2169-35362023-01-0111169841699310.1109/ACCESS.2023.324265410038600CloudNet: A LiDAR-Based Face Anti-Spoofing Model That Is Robust Against Light VariationYongrae Kim0Hyunmin Gwak1Jaehoon Oh2Minho Kang3Jinkyu Kim4https://orcid.org/0000-0001-6520-2074Hyun Kwon5Sunghwan Kim6https://orcid.org/0000-0002-0442-7795Department of Computer Science and Engineering, Korea University, Seoul, South KoreaMustree Company Ltd., Seoul, South KoreaDepartment of Applied Statistics, Konkuk University, Seoul, South KoreaDepartment of Physics, Korea University, Seoul, South KoreaDepartment of Computer Science and Engineering, Korea University, Seoul, South KoreaDepartment of Artificial Intelligence and Data Science, Korea Military Academy, Seoul, South KoreaDepartment of Applied Statistics, Konkuk University, Seoul, South KoreaFace anti-spoofing (FAS) is a technology that protects face recognition systems from presentation attacks. The current challenge faced by FAS studies is the difficulty in creating a generalized light variation model. This is because face data are sensitive to light domain. FAS models using only red green blue (RGB) images suffer from poor performance when the training and test datasets have different light variations. To overcome this problem, this study focuses on light detection and ranging (LiDAR) sensors. LiDAR is a time-of-flight depth sensor that is included in the latest mobile devices. It is negligibly affected by light and provides 3D coordinate and depth information of the target. Thus, a model that is resistant to light variations and exhibiting excellent performance can be created. For the experiment, datasets collected with a LiDAR camera are built and CloudNet architectures for RGB, point clouds, and depth are designed. Three protocols are used to confirm the performance of the model according to variations in the light domain. Experimental results indicate that for protocols 2 and 3, CloudNet error rates increase by 0.1340 and 0.1528, whereas the error rates of the RGB model increase by 0.3951 and 0.4111, respectively, as compared with protocol 1. These results demonstrate that the LiDAR-based FAS model with CloudNet has a more generalized performance compared with the RGB model.https://ieeexplore.ieee.org/document/10038600/Deep learningface anti-spoofingLiDARpoint cloud |
| spellingShingle | Yongrae Kim Hyunmin Gwak Jaehoon Oh Minho Kang Jinkyu Kim Hyun Kwon Sunghwan Kim CloudNet: A LiDAR-Based Face Anti-Spoofing Model That Is Robust Against Light Variation IEEE Access Deep learning face anti-spoofing LiDAR point cloud |
| title | CloudNet: A LiDAR-Based Face Anti-Spoofing Model That Is Robust Against Light Variation |
| title_full | CloudNet: A LiDAR-Based Face Anti-Spoofing Model That Is Robust Against Light Variation |
| title_fullStr | CloudNet: A LiDAR-Based Face Anti-Spoofing Model That Is Robust Against Light Variation |
| title_full_unstemmed | CloudNet: A LiDAR-Based Face Anti-Spoofing Model That Is Robust Against Light Variation |
| title_short | CloudNet: A LiDAR-Based Face Anti-Spoofing Model That Is Robust Against Light Variation |
| title_sort | cloudnet a lidar based face anti spoofing model that is robust against light variation |
| topic | Deep learning face anti-spoofing LiDAR point cloud |
| url | https://ieeexplore.ieee.org/document/10038600/ |
| work_keys_str_mv | AT yongraekim cloudnetalidarbasedfaceantispoofingmodelthatisrobustagainstlightvariation AT hyunmingwak cloudnetalidarbasedfaceantispoofingmodelthatisrobustagainstlightvariation AT jaehoonoh cloudnetalidarbasedfaceantispoofingmodelthatisrobustagainstlightvariation AT minhokang cloudnetalidarbasedfaceantispoofingmodelthatisrobustagainstlightvariation AT jinkyukim cloudnetalidarbasedfaceantispoofingmodelthatisrobustagainstlightvariation AT hyunkwon cloudnetalidarbasedfaceantispoofingmodelthatisrobustagainstlightvariation AT sunghwankim cloudnetalidarbasedfaceantispoofingmodelthatisrobustagainstlightvariation |