Model by nvidia

Studio Voice

Enhance input speech recorded with low-quality microphones in noisy or reverberant environments, producing studio-quality speech.