Advancing Zero-shot Text-to-Speech Intelligibility across Diverse Domains via Preference Alignment

Re-assign community