AI Powered Deepfake Voice and Scam Call Detector for Secure Communication

DOI

10.2991/978-94-6463-948-3_49How to use a DOI?

Keywords

Deepfake Voice Detection,CNN-RNN Hybrid model,Spectrograms; Digital Forensics; Explainable AI

Abstract

The voices synthesized by AI, as well as deepfake audio applications, are a significant threat to identification software and the credibility of an individual, allowing fraud, deception, and fake news. This is getting more difficult to detect fake voices as synthetic speech gets more realistic.The proposed deep learning system in this paper employs both Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) in the identification of an AI-generated voice as well as voices that have been manipulated. It has been trained on a balanced data of real and synthetic voice records. The time and frequency-domain characteristics are generated using Linux audio signal processing environment with spectrogram analysis and Mel-frequency Cepstro Coefficients (MFCC) feature extraction. Additionally, Explainable AI (XAI) modules allow it to make the decision clear and understandable.The results of experimental analysis reveal that a hybrid CNN-RNN is highly accurate and discriminatory when distinguishing real and fake voices and has low error rates. XAI also makes the models more reliable as it offers insight into the outcome of the prediction process.The results are positive towards the practicability of the described approach in telecommunication systems, digital forensics, and fraud detection. The research will help develop safe, credible, and explainable solutions to overcome the threat posed by deepfake audio.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Madhvi Saxena
AU  - Omkar Pote
AU  - Varad Satarkar
AU  - Yash Bhosale
AU  - Soham Ranjane
AU  - Gayatri Badkar
PY  - 2026
DA  - 2026/01/06
TI  - AI Powered Deepfake Voice and Scam Call Detector for Secure Communication
BT  - Proceedings of the International Conference on Sustainable Innovation with Artificial Intelligence and Machine Learning 2025 (ICSIAIML 2025)
PB  - Atlantis Press
SP  - 705
EP  - 719
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6463-948-3_49
DO  - 10.2991/978-94-6463-948-3_49
ID  - Saxena2026
ER  -