Speech Enhancement for Headphones Based on Spectrum Subtraction and DMA

Jiaqi Yu
Proceedings of 2025 2nd International Conference on Machine Learning and Intelligent Computing, PMLR 278:57-66, 2025.

Abstract

For headphone speech enhancement, traditional mono noise reduction methods, such as Wiener filter, have limited noise reduction effects and large computational costs, the directivity of the differential microphone array (DMA) is used to optimize the voice pickup and improve the signal-to-noise ratio of the input signal. Experimental simulation results show that the proposed method improves the output signal-to-noise ratio by 6dB under different output signal-to-noise ratios, and still has a 2dB speech enhancement effect under the condition of low signal-to-noise ratio of -5dB, and has strong anti-interference ability. The effect is stable in the background of broadband noise and low-frequency noise, which can better optimize the user’s experience during call.

Cite this Paper

BibTeX
@InProceedings{pmlr-v278-yu25a, title = {Speech Enhancement for Headphones Based on Spectrum Subtraction and DMA}, author = {Yu, Jiaqi}, booktitle = {Proceedings of 2025 2nd International Conference on Machine Learning and Intelligent Computing}, pages = {57--66}, year = {2025}, editor = {Zeng, Nianyin and Pachori, Ram Bilas and Wang, Dongshu}, volume = {278}, series = {Proceedings of Machine Learning Research}, month = {25--27 Apr}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v278/main/assets/yu25a/yu25a.pdf}, url = {https://proceedings.mlr.press/v278/yu25a.html}, abstract = {For headphone speech enhancement, traditional mono noise reduction methods, such as Wiener filter, have limited noise reduction effects and large computational costs, the directivity of the differential microphone array (DMA) is used to optimize the voice pickup and improve the signal-to-noise ratio of the input signal. Experimental simulation results show that the proposed method improves the output signal-to-noise ratio by 6dB under different output signal-to-noise ratios, and still has a 2dB speech enhancement effect under the condition of low signal-to-noise ratio of -5dB, and has strong anti-interference ability. The effect is stable in the background of broadband noise and low-frequency noise, which can better optimize the user’s experience during call.} }
Endnote
%0 Conference Paper %T Speech Enhancement for Headphones Based on Spectrum Subtraction and DMA %A Jiaqi Yu %B Proceedings of 2025 2nd International Conference on Machine Learning and Intelligent Computing %C Proceedings of Machine Learning Research %D 2025 %E Nianyin Zeng %E Ram Bilas Pachori %E Dongshu Wang %F pmlr-v278-yu25a %I PMLR %P 57--66 %U https://proceedings.mlr.press/v278/yu25a.html %V 278 %X For headphone speech enhancement, traditional mono noise reduction methods, such as Wiener filter, have limited noise reduction effects and large computational costs, the directivity of the differential microphone array (DMA) is used to optimize the voice pickup and improve the signal-to-noise ratio of the input signal. Experimental simulation results show that the proposed method improves the output signal-to-noise ratio by 6dB under different output signal-to-noise ratios, and still has a 2dB speech enhancement effect under the condition of low signal-to-noise ratio of -5dB, and has strong anti-interference ability. The effect is stable in the background of broadband noise and low-frequency noise, which can better optimize the user’s experience during call.
APA
Yu, J.. (2025). Speech Enhancement for Headphones Based on Spectrum Subtraction and DMA. Proceedings of 2025 2nd International Conference on Machine Learning and Intelligent Computing, in Proceedings of Machine Learning Research 278:57-66 Available from https://proceedings.mlr.press/v278/yu25a.html.

Related Material