Audio Deepfake Detector

Objectives

• Upload or record audio samples for authenticity analysis
• Extract spectral features like MFCCs, spectrograms, and formants
• Detect common AI-generated audio artifacts and anomalies
• Classify audio as real or synthetic with a confidence score
• Display a detailed breakdown of detection signals and visualizations

Hints

• AI-generated audio often has subtle spectral smoothing artifacts
• Use a pre-trained model or build a simple classifier with labeled data
• Display spectrograms to visually highlight differences between real and fake audio

Resources

• developer.mozilla.org/en-US/docs/Web/API/Web_Audio_API

Comments (0)