CoNSIST: Consist of New Methodologies on AASIST for Audio Deepfake Detection
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Advancements in artificial intelligence(AI) have significantly improved deep learning-based audio deepfake technology, which has beenexploited for criminal activities. To detect audio deepfake, we propose CoNSIST, an advanced audio deepfake detection model. CoNSISTbuilds on AASIST, which a graph-based end-to-end model, by integrating three key components: Squeeze and Excitation, PositionalEncoding, and Reformulated HS-GAL. These additions aim to enhance feature extraction, eliminate unnecessary operations, andincorporate diverse information. Our experimental results demonstrate that CoNSIST significantly outperforms existing models in detectingaudio deepfakes, offering a more robust solution to combat the misuse of this technology
Abstract
Advancements in artificial intelligence(AI) have significantly improved deep learning-based audio deepfake technology, which has beenexploited for criminal activities. To detect audio deepfake, we propose CoNSIST, an advanced audio deepfake detection model. CoNSISTbuilds on AASIST, which a graph-based end-to-end model, by integrating three key components: Squeeze and Excitation, PositionalEncoding, and Reformulated HS-GAL. These additions aim to enhance feature extraction, eliminate unnecessary operations, andincorporate diverse information. Our experimental results demonstrate that CoNSIST significantly outperforms existing models in detectingaudio deepfakes, offering a more robust solution to combat the misuse of this technology
참고문헌 (21)
- "Audio Deepfake Detection: A Survey" (2023)[URL]
- J. W. Jung, "AASIST: Audio Anti-Spoofing Using Integrated Spectro-Temporal Graph Attention Networks", ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) (2022)
- A. Hamza, "Deepfake Audio Detection via MFCC Features Using Machine Learning", IEEE Access, 10, pp.134018~134028 (2022)[DOI]
- M. Lataifeh, "Ar-DAD: Arabic Diversified Audio Dataset", Data in Brief, pp.106503~ (2020)[DOI]
- C. Borrelli, "Synthetic Speech Detection through Short-Term and Long-Term Prediction Traces", EURASIP Journal on Information Security, (1), pp.1186~ (2021)[DOI]
- A. K. Singh, "Detection of AI-Synthesized Speech Using Cepstral & Bispectral Statistics", 2021 IEEE 4th International Conference on Multimedia Information Processing and Retrieval (MIPR) (2021)
- A. Chintha, "Recurrent Convolutional Structures for Audio Spoof and Video Deepfake Detection", IEEE Journal of Selected Topics in Signal Processing, 14(5), pp.1024~1037 (2020)[DOI]
- X. Liu, "Leveraging Positional-Related Local-Global Dependency for Synthetic Speech Detection" (2023)[DOI]
- "End-To-End Spectro-Temporal Graph Attention Networks for Speaker Verification Anti-Spoofing and Speech Deepfake Detection" (2021)[URL]
- "Graph Attention Networks for Anti-Spoofing" (2021)[URL]
- "Improved RawNet with Feature Map Scaling for Text-Independent Speaker Verification Using Raw Waveforms" (2020)[URL]
- H. Tak, "End-To-End Anti-Spoofing with RawNet2" (2021), HAL (Le Centre Pour La Communication Scientifique Directe)
- X. Wang, "Heterogeneous Graph Attention Network", The World Wide Web Conference (2019)
- "Graph Attention Networks" (2018)[URL]
- J. Hu, "Squeeze-andExcitation Networks", 2018 IEEE, 2018, ~
- P. Duffer, "Position Information in Transformers : An Overview", Computational Linguistics, 48(3), pp.733~763 (2022)[DOI]
- "Attention is All you Need" (2017)[URL]
- "ASVspoof 2019: A Large-Scale Public Database of Synthesized, Converted and Replayed Speech" (2019)[URL]
- T. Kinnunen, "T-DCF: A Detection Cost Function for the Tandem Assessment of Spoofing Countermeasures and Automatic Speaker Verification", Odyssey 2018 the Speaker and Language Recognition Workshop (2018)
- "A Comparative Study on Recent Neural Spoofing Countermeasures for Synthetic Speech Detection" (2021)[URL]