Three-Dimensional Convolutional Approaches for the Verification of Deepfake Videos: The Effect of Image Depth Size on Authentication Performance
Deep learning has proven to be particularly effective in tasks such as data analysis, computer vision, and human control. However, as this method has become more advanced, it has also led to the creation of DeepFake video sequences and images in which alterations can be made without immediately appe...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Engineering and Technology Publishing
2023
|
| Subjects: | |
| Online Access: | http://umpir.ump.edu.my/id/eprint/39131/ http://umpir.ump.edu.my/id/eprint/39131/1/4.May_JAIT-V14N3-488.pdf |
| _version_ | 1848825691935604736 |
|---|---|
| author | Muhammad Salihin, Saealal Mohd Zamri, Ibrahim Marlina, Yakno Nurul Wahidah, Arshad |
| author_facet | Muhammad Salihin, Saealal Mohd Zamri, Ibrahim Marlina, Yakno Nurul Wahidah, Arshad |
| author_sort | Muhammad Salihin, Saealal |
| building | UMP Institutional Repository |
| collection | Online Access |
| description | Deep learning has proven to be particularly effective in tasks such as data analysis, computer vision, and human control. However, as this method has become more advanced, it has also led to the creation of DeepFake video sequences and images in which alterations can be made without immediately appealing to the viewer. These technological advancements have introduced new security threats, including in the field of education. For example, in online exams and tests conducted through video conferencing, individuals may use Deepfake technology to impersonate another person, potentially allowing them to cheat by having someone else take the exam in their place. Several detection approaches have been proposed to address these issues, including systems that use both spatial and temporal features. However, existing approaches have limitations regarding detection accuracy and overall effectiveness. The paper proposes a technique for detecting Deepfakes that combines temporal analysis with convolutional neural networks. The study explores various 3-D Convolutional Neural Networks-based (CNN-based) model approaches and different sequence lengths of facial photos. The results indicate that using a 3-D CNN model with 16 sequential face images as input can detect Deepfakes with up to 97.3 percent accuracy on the FaceForensic dataset. Detecting Deepfakes is crucial as they pose a threat to the authenticity of visual media. The proposed technique offers a promising solution to this issue. |
| first_indexed | 2025-11-15T03:32:57Z |
| format | Article |
| id | ump-39131 |
| institution | Universiti Malaysia Pahang |
| institution_category | Local University |
| language | English |
| last_indexed | 2025-11-15T03:32:57Z |
| publishDate | 2023 |
| publisher | Engineering and Technology Publishing |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | ump-391312023-11-01T08:11:35Z http://umpir.ump.edu.my/id/eprint/39131/ Three-Dimensional Convolutional Approaches for the Verification of Deepfake Videos: The Effect of Image Depth Size on Authentication Performance Muhammad Salihin, Saealal Mohd Zamri, Ibrahim Marlina, Yakno Nurul Wahidah, Arshad Q Science (General) QA75 Electronic computers. Computer science Deep learning has proven to be particularly effective in tasks such as data analysis, computer vision, and human control. However, as this method has become more advanced, it has also led to the creation of DeepFake video sequences and images in which alterations can be made without immediately appealing to the viewer. These technological advancements have introduced new security threats, including in the field of education. For example, in online exams and tests conducted through video conferencing, individuals may use Deepfake technology to impersonate another person, potentially allowing them to cheat by having someone else take the exam in their place. Several detection approaches have been proposed to address these issues, including systems that use both spatial and temporal features. However, existing approaches have limitations regarding detection accuracy and overall effectiveness. The paper proposes a technique for detecting Deepfakes that combines temporal analysis with convolutional neural networks. The study explores various 3-D Convolutional Neural Networks-based (CNN-based) model approaches and different sequence lengths of facial photos. The results indicate that using a 3-D CNN model with 16 sequential face images as input can detect Deepfakes with up to 97.3 percent accuracy on the FaceForensic dataset. Detecting Deepfakes is crucial as they pose a threat to the authenticity of visual media. The proposed technique offers a promising solution to this issue. Engineering and Technology Publishing 2023 Article PeerReviewed pdf en cc_by_4 http://umpir.ump.edu.my/id/eprint/39131/1/4.May_JAIT-V14N3-488.pdf Muhammad Salihin, Saealal and Mohd Zamri, Ibrahim and Marlina, Yakno and Nurul Wahidah, Arshad (2023) Three-Dimensional Convolutional Approaches for the Verification of Deepfake Videos: The Effect of Image Depth Size on Authentication Performance. Journal of Advances in Information Technology, 14 (3). pp. 488-494. ISSN 1798-2340. (Published) https://doi.org/10.12720/jait.14.3.488-494 10.12720/jait.14.3.488-494 |
| spellingShingle | Q Science (General) QA75 Electronic computers. Computer science Muhammad Salihin, Saealal Mohd Zamri, Ibrahim Marlina, Yakno Nurul Wahidah, Arshad Three-Dimensional Convolutional Approaches for the Verification of Deepfake Videos: The Effect of Image Depth Size on Authentication Performance |
| title | Three-Dimensional Convolutional Approaches for the Verification of Deepfake Videos: The Effect of Image Depth Size on Authentication Performance |
| title_full | Three-Dimensional Convolutional Approaches for the Verification of Deepfake Videos: The Effect of Image Depth Size on Authentication Performance |
| title_fullStr | Three-Dimensional Convolutional Approaches for the Verification of Deepfake Videos: The Effect of Image Depth Size on Authentication Performance |
| title_full_unstemmed | Three-Dimensional Convolutional Approaches for the Verification of Deepfake Videos: The Effect of Image Depth Size on Authentication Performance |
| title_short | Three-Dimensional Convolutional Approaches for the Verification of Deepfake Videos: The Effect of Image Depth Size on Authentication Performance |
| title_sort | three-dimensional convolutional approaches for the verification of deepfake videos: the effect of image depth size on authentication performance |
| topic | Q Science (General) QA75 Electronic computers. Computer science |
| url | http://umpir.ump.edu.my/id/eprint/39131/ http://umpir.ump.edu.my/id/eprint/39131/ http://umpir.ump.edu.my/id/eprint/39131/ http://umpir.ump.edu.my/id/eprint/39131/1/4.May_JAIT-V14N3-488.pdf |