CoolFace
Apppublic

KaushiGihan/Deepfake_audio_detection_yamnet

sourceHugging Facemitupdated 2y agoView on Hugging Face
1likes
App README

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

Deepfake-voicedetectionYamnet

This repository contains a machine learning-based system to detect deepfake (synthetic) voices. The system utilizes audio feature extraction techniques such as YAMNet and deep learning models (ANN, CNN, RNN) to differentiate between real and fake audio.

🚀 Features

  • —Audio Preprocessing: Converts raw audio into meaningful features.
  • —Feature Extraction: Uses YAMNet to extract embeddings from audio signals.
  • —Dataset : Uses the kaggle "In The Wild (audio Deepfake)" dataset for training the model.Link
  • —Deep Learning Models: Implements ANN, CNN, and RNN architectures for classification.
  • —Training & Evaluation: Trains models with labeled datasets and evaluates accuracy.
  • —Inference API: Provides an API to classify input audio as real or deepfake.

Model Performance

  1. 1.ANN model atchitecture
  2. 2.Evaluation Metrics: Accuracy, Precision, Recall, F1-score

precision recall f1-score support fake 0.97 0.95 0.96 446 real 0.96 0.98 0.97 554

accuracy 0.97 1000 macro avg 0.97 0.97 0.97 1000 weighted avg 0.97 0.97 0.97 1000

  1. 1.CNN model atchitecture
  2. 2.Evaluation Metrics: Accuracy, Precision, Recall, F1-score

precision recall f1-score support fake 0.97 0.96 0.97 446 real 0.97 0.97 0.97 554

accuracy 0.97 1000 macro avg 0.97 0.97 0.97 1000 weighted avg 0.97 0.97 0.97 1000

  1. 1.RNN model atchitecture
  • —Evaluation Metrics: Accuracy, Precision, Recall, F1-score

precision recall f1-score support fake 0.94 0.93 0.94 446 real 0.94 0.95 0.95 554

accuracy 0.94 1000 macro avg 0.94 0.94 0.94 1000 weighted avg 0.94 0.94 0.94 1000

Best Model Achieved: ANN with 97% accuracy

🛠️ Setup & Installation

1.Clone the repository:

python
git clone https://github.com/KaushiML3/Deepfake-voice-detection_Yamnet.git
cd Deepfake-voice-detection_Yamnet

2.Install dependencies:

python
pip install -r requirements.txt

3.Run inference

  • —change the direction for API folder
python
python main.py