Audio Deepfake Detection is a web page that uses machine learning techniques to analyze audio files and determine if they are real or generated by deepfake algorithms.
- Audio feature extraction using MFCC, mel spectrogram, chroma features, zero crossing rate, spectral centroid, and spectral flatness.
- Comparison of extracted features with a pre-defined dataset.
- Classification of audio as either real or deepfake.
- Visual representation of the analysis results.
1.Upload an audio file for analysis. 2.Wait for the analysis to complete. 3.View the result, indicating whether the uploaded audio is classified as real or a deepfake. 4.Repeat the process for additional audio files.
Contributions are welcome! If you find any issues or have suggestions for improvements, please submit a pull request or open an issue in the GitHub repository.
This project was developed as part of a college project and should not be considered as a foolproof solution for deepfake detection. While efforts have been made to provide accurate results, the system has limitations and may not be fully effective in all scenarios. It is recommended to use multiple layers of security and verification methods for critical applications. For any important or high-stakes scenarios, consult with experts and consider employing more advanced and specialized deepfake detection techniques.
To enhance the accuracy of this deepfake detection system, your contribution is valuable. If you have additional deepfake audio samples or expertise in deepfake detection, we welcome your collaboration and encourage you to contribute to the project. By expanding the dataset and incorporating more diverse samples, we can collectively improve the effectiveness of the system. Please feel free to reach out and contribute to this ongoing effort.
For inquiries, please contact [jeraldgolden00@gmail.com].