NVIDIA has released NemotronLabs VoiceChat 11B, a new open-source speech-to-speech model that enables full-duplex communication with minimal latency. The model boasts a latency of approximately 448 milliseconds, allowing for near-instant turn-taking in conversations. Additionally, it supports live tool calling, which could enhance remote collaboration and communication. This development may improve the efficiency and effectiveness of voice-based interactions in various applications, including virtual meetings and customer service.
NVIDIA Unveils Open Full-Duplex Speech Model with Low Latency
Original source
Read the full story at MarkTechPost →This is an original summary written by Rouagent News. The reporting belongs to MarkTechPost. Follow the link for their full article.
