October 20, 2024: Metas Spirit LM Enhances Expressive AI Voices - Meta Platforms Inc. introduces Spirit LM, an open-source multimodal AI model designed to generate expressive, emotion-rich voices. Unlike traditional models, Spirit LM incorporates phonetic, pitch, and tone tokens to produce more human-like speech. It can learn tasks across modalities, including text-to-speech and speech classification. Two versions, Spirit LM Base and Spirit LM Expressive, are available for noncommercial research, with the latter capable of replicating emotions like anger and happiness. Aimed at improving human-machine interaction, Spirit LM is part of Metas broader initiative to explore advanced machine intelligence.
