Improving Video Vision Transformer for Deepfake Video Detection Using Facial Landmark, Depthwise Separable Convolution and Self Attention
In this paper, we present our result of research in video deepfake detection. We built a deepfake detection system to detect whether a video is a deepfake or real. The deepfake detection algorithm still struggle in providing a sufficient accuracy values, especially in challenging deepfake dataset. O...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10388363/ |
| _version_ | 1797348686378827776 |
|---|---|
| author | Kurniawan Nur Ramadhani Rinaldi Munir Nugraha Priya Utama |
| author_facet | Kurniawan Nur Ramadhani Rinaldi Munir Nugraha Priya Utama |
| author_sort | Kurniawan Nur Ramadhani |
| collection | DOAJ |
| description | In this paper, we present our result of research in video deepfake detection. We built a deepfake detection system to detect whether a video is a deepfake or real. The deepfake detection algorithm still struggle in providing a sufficient accuracy values, especially in challenging deepfake dataset. Our deepfake detection system utilized spatiotemporal feature that extracted using Video Vision Transformer (ViViT). The main contribution of our research is providing a deepfake detection system that based on ViViT architecture and using landmark area images for the input of the system. Our system extracted the feature from a number of spatial features. The spatial feature was extracted using Depthwise Separable Convolution (DSC) block combined with Convolution Block Attention Module (CBAM) from tubelet. The tubelet was a representation of facial landmark area that was extracted from the input video. In our system, we used 25 facial landmark area for an input video. In our experiment we used Celeb-DF version 2 dataset because it is considered to be a challenging deepfake dataset. We conducted augmentation to the dataset, so we obtained 8335 videos for training set, 390 videos for validation set, and 1123 videos for testing set. We trained our deepfake detection system using Adam optimizer, with learning rate of 10–4 and 100 epoch. From the experiment, we obtained the accuracy score of 87.18% and F1 score of 92.52%. We also conducted the ablation study to display the effect of each part of our model to the overall system performance. From this research, we obtained that by using landmark area images, our ViViT based deepfake detection system had a good performance in detecting deepfake videos. |
| first_indexed | 2024-03-08T12:09:33Z |
| format | Article |
| id | doaj.art-4b28f1b0d3694d048f51c9723456b820 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-03-08T12:09:33Z |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-4b28f1b0d3694d048f51c9723456b8202024-01-23T00:05:51ZengIEEEIEEE Access2169-35362024-01-01128932893910.1109/ACCESS.2024.335289010388363Improving Video Vision Transformer for Deepfake Video Detection Using Facial Landmark, Depthwise Separable Convolution and Self AttentionKurniawan Nur Ramadhani0https://orcid.org/0000-0002-5126-8213Rinaldi Munir1Nugraha Priya Utama2Bandung Institute of Technology, Bandung, IndonesiaBandung Institute of Technology, Bandung, IndonesiaBandung Institute of Technology, Bandung, IndonesiaIn this paper, we present our result of research in video deepfake detection. We built a deepfake detection system to detect whether a video is a deepfake or real. The deepfake detection algorithm still struggle in providing a sufficient accuracy values, especially in challenging deepfake dataset. Our deepfake detection system utilized spatiotemporal feature that extracted using Video Vision Transformer (ViViT). The main contribution of our research is providing a deepfake detection system that based on ViViT architecture and using landmark area images for the input of the system. Our system extracted the feature from a number of spatial features. The spatial feature was extracted using Depthwise Separable Convolution (DSC) block combined with Convolution Block Attention Module (CBAM) from tubelet. The tubelet was a representation of facial landmark area that was extracted from the input video. In our system, we used 25 facial landmark area for an input video. In our experiment we used Celeb-DF version 2 dataset because it is considered to be a challenging deepfake dataset. We conducted augmentation to the dataset, so we obtained 8335 videos for training set, 390 videos for validation set, and 1123 videos for testing set. We trained our deepfake detection system using Adam optimizer, with learning rate of 10–4 and 100 epoch. From the experiment, we obtained the accuracy score of 87.18% and F1 score of 92.52%. We also conducted the ablation study to display the effect of each part of our model to the overall system performance. From this research, we obtained that by using landmark area images, our ViViT based deepfake detection system had a good performance in detecting deepfake videos.https://ieeexplore.ieee.org/document/10388363/Deepfake detectionfacial landmarkdepthwise separable convolutionconvolution block attention modulevideo vision transformer |
| spellingShingle | Kurniawan Nur Ramadhani Rinaldi Munir Nugraha Priya Utama Improving Video Vision Transformer for Deepfake Video Detection Using Facial Landmark, Depthwise Separable Convolution and Self Attention IEEE Access Deepfake detection facial landmark depthwise separable convolution convolution block attention module video vision transformer |
| title | Improving Video Vision Transformer for Deepfake Video Detection Using Facial Landmark, Depthwise Separable Convolution and Self Attention |
| title_full | Improving Video Vision Transformer for Deepfake Video Detection Using Facial Landmark, Depthwise Separable Convolution and Self Attention |
| title_fullStr | Improving Video Vision Transformer for Deepfake Video Detection Using Facial Landmark, Depthwise Separable Convolution and Self Attention |
| title_full_unstemmed | Improving Video Vision Transformer for Deepfake Video Detection Using Facial Landmark, Depthwise Separable Convolution and Self Attention |
| title_short | Improving Video Vision Transformer for Deepfake Video Detection Using Facial Landmark, Depthwise Separable Convolution and Self Attention |
| title_sort | improving video vision transformer for deepfake video detection using facial landmark depthwise separable convolution and self attention |
| topic | Deepfake detection facial landmark depthwise separable convolution convolution block attention module video vision transformer |
| url | https://ieeexplore.ieee.org/document/10388363/ |
| work_keys_str_mv | AT kurniawannurramadhani improvingvideovisiontransformerfordeepfakevideodetectionusingfaciallandmarkdepthwiseseparableconvolutionandselfattention AT rinaldimunir improvingvideovisiontransformerfordeepfakevideodetectionusingfaciallandmarkdepthwiseseparableconvolutionandselfattention AT nugrahapriyautama improvingvideovisiontransformerfordeepfakevideodetectionusingfaciallandmarkdepthwiseseparableconvolutionandselfattention |