This repository is related to our Dataset and Detection code from the paper: AI-Synthesized Voice Detection Using Neural Vocoder Artifacts accepted in CVPR Workshop on Media Forensic 2023.

Implementation of the paper "Improved DeepFake Detection Using Whisper Features"

SUTD 50.039 Deep Learning Course Project (2022 Spring)

Implementation of Attack Agnostic Dataset: Towards Generalization and Stabilization of Audio DeepFake Detection paper

Implementation of "Defense against Adversarial Attacks on Audio DeepFake Detection"

Audio deepfake detection sytem on CNN

Official implementation of the INTERSPEECH 2024 paper: Temporal-Channel Modeling in Multi-head Self-Attention for Synthetic Speech Detection

Implementation of "SpecRNet: Towards Faster and More Accessible Audio DeepFake Detection" paper

Baselines for IS25 Source Tracing Special Session

LibriVoc is a new open-source, large-scale dataset for vocoder artifact detection. LibriVoc is derived from the LibriTTS speech corpus, which is widely used in text-to- speech research. The LibriTTS corpus is derived from the Librispeech dataset, wherein each sample is extracted from LibriVox audiobooks.

Application that detects the authenticity of audio files developed using the Random Forest Model.

Audio Deepfake Detection is a web page that utilizes machine learning techniques to analyze audio files and determine if they are real or generated by deepfake algorithms. It features user registration, audio file upload, audio feature extraction, comparison with a pre-defined dataset, and classification of audio as real or deepfake.

Here is the Asvspoof19 Laundered Dataset an deepfake audio website for downloading the data and preview the sample data

This project aims to detect audio deepfakes using a hybrid approach that combines CNN and BiLSTM. The system is designed to effectively classify audio data into genuine or fake categories, offering a robust solution to the growing challenges posed by audio-based misinformation.