Towards efficient and effective face forgery detection
Rapid progress in Face Forgery algorithms is becoming an increasingly relevant threat in the modern-day. To address the issue, multiple state-of-the-art forgery detection algorithms have been proposed. However, these models made use of deep and complex Convolutional Neural Networks (CNN) as their ba...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Final Year Project (FYP) |
| Language: | English |
| Published: |
Nanyang Technological University
2022
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/162785 |
| _version_ | 1811696219644231680 |
|---|---|
| author | Peng, Weixing |
| author2 | Lin Guosheng |
| author_facet | Lin Guosheng Peng, Weixing |
| author_sort | Peng, Weixing |
| collection | NTU |
| description | Rapid progress in Face Forgery algorithms is becoming an increasingly relevant threat in the modern-day. To address the issue, multiple state-of-the-art forgery detection algorithms have been proposed. However, these models made use of deep and complex Convolutional Neural Networks (CNN) as their backbone, making them unrealistic to use in a production environment due to the high requirement for computational resources. Luckily, research on network pruning techniques has made tremendous progress in recent years, making it possible to compress CNN models by 90%, while retaining the same model accuracy. In this project, we experimented with and evaluated multiple pruning strategies and techniques. We then applied our findings by pruning a baseline face forgery detection model and achieved an impressive 2x speedup in the model inference time and a 70% decrease in model size. When our model is run under CPU-only hardware, a 30x speedup was achieved. |
| first_indexed | 2024-10-01T07:35:53Z |
| format | Final Year Project (FYP) |
| id | ntu-10356/162785 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T07:35:53Z |
| publishDate | 2022 |
| publisher | Nanyang Technological University |
| record_format | dspace |
| spelling | ntu-10356/1627852022-11-09T02:16:00Z Towards efficient and effective face forgery detection Peng, Weixing Lin Guosheng School of Computer Science and Engineering gslin@ntu.edu.sg Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision Rapid progress in Face Forgery algorithms is becoming an increasingly relevant threat in the modern-day. To address the issue, multiple state-of-the-art forgery detection algorithms have been proposed. However, these models made use of deep and complex Convolutional Neural Networks (CNN) as their backbone, making them unrealistic to use in a production environment due to the high requirement for computational resources. Luckily, research on network pruning techniques has made tremendous progress in recent years, making it possible to compress CNN models by 90%, while retaining the same model accuracy. In this project, we experimented with and evaluated multiple pruning strategies and techniques. We then applied our findings by pruning a baseline face forgery detection model and achieved an impressive 2x speedup in the model inference time and a 70% decrease in model size. When our model is run under CPU-only hardware, a 30x speedup was achieved. Bachelor of Engineering (Computer Science) 2022-11-09T02:16:00Z 2022-11-09T02:16:00Z 2022 Final Year Project (FYP) Peng, W. (2022). Towards efficient and effective face forgery detection. Final Year Project (FYP), Nanyang Technological University, Singapore. https://hdl.handle.net/10356/162785 https://hdl.handle.net/10356/162785 en SCSE21-0635 application/pdf Nanyang Technological University |
| spellingShingle | Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision Peng, Weixing Towards efficient and effective face forgery detection |
| title | Towards efficient and effective face forgery detection |
| title_full | Towards efficient and effective face forgery detection |
| title_fullStr | Towards efficient and effective face forgery detection |
| title_full_unstemmed | Towards efficient and effective face forgery detection |
| title_short | Towards efficient and effective face forgery detection |
| title_sort | towards efficient and effective face forgery detection |
| topic | Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision |
| url | https://hdl.handle.net/10356/162785 |
| work_keys_str_mv | AT pengweixing towardsefficientandeffectivefaceforgerydetection |