Disentangling Emotion from Voice: A Cross- Sampling Approach for Expressive Voice Data

Hume· 17 SepResearch paper

Introducing the Hume Voice Controllability Leaderboard

Leaderboard tests TTS model instruction following

Hume· 10 Sep

Introducing the Hume Voice Replication Leaderboard

New leaderboard ranks eleven text-to-speech models on voice replication quality.

Hume· 21 AugResearch paper

Measuring benchmark optimization in speech recognition