L. Moro-Velázquez

ProPS can generate speaker embeddings that accurately reflect complex attributes from simple natural language prompts, revolutionizing how we synthesize speaker identities.

Current speech-to-speech models may sound good, but they miss the mark on natural conversational dynamics, revealing critical areas for improvement.

Emotional entrainment detection can reach 97.01% accuracy by considering the temporal dynamics of dyadic speech interactions.

Speech-aware LLMs are surprisingly bad at speaker verification, but a simple embedding injection trick closes the gap with dedicated systems while preserving the LLM's language abilities.