Phonetic Fidelity Voice Conversion evaluates how well voice conversion models preserve phonetic details by comparing source and converted speech using Phoneme Error Rate (PER), complementing traditional text-based metrics like WER and CER

This Project presents 11Labs Clone, a comprehensive AI voice generation platform that combines multiple stateof-the-art text-to-speech and voice cloning technologies. The platform integrates StyleTTS2 for high-quality text-tospeech synthesis, SEED-VC for voice cloning capabilities, and Make-An-Audio for advanced audio generation.