Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces a low-latency deep learning model for real-time vocal denoising using a band-grouped encoder-decoder architecture with frequency attention. The model is trained with a spectral loss and sigmoid-driven ideal ratio mask to improve SNR and perceptual quality. Results show the model achieves less than 10ms latency with PESQ-WB improvements of 0.21 and 0.12 on stationary and non-stationary noise, respectively.
Real-time vocal denoising is now possible with deep learning, achieving significant SNR improvements at under 10ms latency.
Real-time, deep learning-based vocal denoising has seen significant progress over the past few years, demonstrating the capability of artificial intelligence in preserving the naturalness of the voice while increasing the signal-to-noise ratio (SNR). However, many deep learning approaches have high amounts of latency and require long frames of context, making them difficult to configure for live applications. To address these challenges, we propose a sigmoid-driven ideal ratio mask trained with a spectral loss to encourage an increased SNR and maximized perceptual quality of the voice. The proposed model uses a band-grouped encoder-decoder architecture with frequency attention and achieves a total latency of less than 10,ms, with PESQ-WB improvements of 0.21 on stationary noise and 0.12 on nonstationary noise.