CheapNET: Improving Light-weight speech enhancement network by projected loss function

TL;DR

CheapNET introduces a novel projection loss function for lightweight speech enhancement models, significantly improving noise suppression and echo cancellation performance while maintaining low computational complexity for real-time applications.

Contribution

The paper proposes a new projection loss function that enhances noise suppression and echo cancellation in lightweight speech enhancement networks, surpassing traditional MSE-based methods.

Findings

Achieves near state-of-the-art noise suppression with 3.1M parameters.

Outperforms industry-leading echo cancellation models.

Maintains low computational load suitable for edge devices.

Abstract

Noise suppression and echo cancellation are critical in speech enhancement and essential for smart devices and real-time communication. Deployed in voice processing front-ends and edge devices, these algorithms must ensure efficient real-time inference with low computational demands. Traditional edge-based noise suppression often uses MSE-based amplitude spectrum mask training, but this approach has limitations. We introduce a novel projection loss function, diverging from MSE, to enhance noise suppression. This method uses projection techniques to isolate key audio components from noise, significantly improving model performance. For echo cancellation, the function enables direct predictions on LAEC pre-processed outputs, substantially enhancing performance. Our noise suppression model achieves near state-of-the-art results with only 3.1M parameters and 0.4GFlops/s computational load. Moreover, our echo cancellation model outperforms replicated industry-leading models, introducing a new perspective in speech enhancement.