The project presents an SVM model trained to distinguish between Real human voices and AI-generated voices.
-
Updated
Feb 26, 2025 - Python
The project presents an SVM model trained to distinguish between Real human voices and AI-generated voices.
A deep learning web application that analyzes audio frequencies to detect whether a voice is AI-generated or a real human speech.
Detects AI-generated audio using Shannon entropy features + Random Forest. 100% blind test accuracy. No GPU required. Trained on LibriSpeech vs Kokoro TTS.
An HCL Guvi Buildathon project
Local AI voice-over suspicion checker for videos and Telegram bot workflows.
To associate your repository with the ai-voice-detection topic, visit your repo's landing page and select "manage topics."