Announcement_0

🥇 Our team placed first in Track 3 of the VoiceMOS Challenge 2026 (predicting speaker and accent similarity for codec-based speech synthesis).