Abstract: The rise of deepfake technology poses a significant threat to the authenticity and integrity of multimedia content, including audio recordings. In response to this challenge, this project proposes a deep learning-based approach for detecting deepfake audio. Leveraging advancements in machine learning and signal processing, the proposed system aims to distinguish between genuine and manipulated audio recordings with high accuracy.The project begins with a comprehensive exploration of existing deepfake detection techniques, focusing on their limitations and strengths, particularly in the context of audio manipulation. Subsequently, a novel deep learning architecture is designed and implemented to effectively capture the subtle cues and patterns indicative of audio manipulation.Key components of the proposed system include feature extraction modules tailored to the unique characteristics of audio data, as well as deep neural network models trained on large-scale datasets of both genuine and deepfake audio samples. Through extensive experimentation and evaluation, the effectiveness and robustness of the developed system are assessed across various types of audio manipulation techniques and levels of sophistication.
Keywords: Deepfake, Audio manipulation, Deep learning, Detection, Feature extraction, Neural networks
Cite:
Ankith Shetty, Hanzala Karani , Shreya K H, Raheeza Khan, Mr. Amruth A G,"Deepfake Audio Detection using Deep Learning", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 3, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13344.
|
DOI:
10.17148/IJARCCE.2024.13344
