Ultra Low Complexity Deep Learning Based Noise Suppression
This addresses the challenge of deploying effective noise suppression on low-power devices, though it appears incremental as it builds on existing methods with modifications.
The paper tackles the problem of high computational complexity in deep neural networks for real-time speech enhancement on resource-constrained devices, achieving noise suppression performance comparable to state-of-the-art methods with 3 to 4 times less computational complexity and memory usage.
This paper introduces an innovative method for reducing the computational complexity of deep neural networks in real-time speech enhancement on resource-constrained devices. The proposed approach utilizes a two-stage processing framework, employing channelwise feature reorientation to reduce the computational load of convolutional operations. By combining this with a modified power law compression technique for enhanced perceptual quality, this approach achieves noise suppression performance comparable to state-of-the-art methods with significantly less computational requirements. Notably, our algorithm exhibits 3 to 4 times less computational complexity and memory usage than prior state-of-the-art approaches.