PRIME-ANC: Path-Ratio-Informed Modeling for Efficient Neural Filter Synthesis in Active Noise Control

Yaokun Huang; Chunyang Xu; Haowen Hua; Sen Lin; Shichao Hu; Mengyao Zhu

eess.ASSep 24, 2026

PRIME-ANC: Path-Ratio-Informed Modeling for Efficient Neural Filter Synthesis in Active Noise Control

Authors: Yaokun Huang, Chunyang Xu, Haowen Hua, Sen Lin, Shichao Hu, Mengyao Zhu

Organizations: Center of Auditory and Language Intelligence, Soochow University, Suzhou, China · School of Future Science and Engineering, Soochow University, Suzhou, China

Abstract

Changes in listener acoustics require active noise control (ANC) filters to be redesigned for new acoustic paths. We introduce PRIME-ANC, a shared neural synthesizer that learns a bounded, path-dependent logmagnitude correction to a regularized path-ratio base. Minimum-phase reconstruction and truncation produce finite-impulse-response (FIR) filters; training optimizes their noise-control performance. Across ten random training/test splits per dataset, PRIME-ANC achieves average held-out one-third-octave reductions of 18.81 and 17.76 dB over 50 Hz-5 kHz on a ten-path dataset and a public earphone database, respectively. On original earphone measurements, it improves upon the path-ratio base by 7.89 dB. Ablation studies support the contributions of both the analytic base and the path-dependent correction. Given calibrated paths for a held-out listener condition, PRIME-ANC generates a path-specific FIR without iterative optimization. With three Gauss-Newton updates, it reaches 21.43 dB reduction, approaching direct weighted least-squares design while producing lower amplification and root-mean-square control output.

Figures & tables

Figure 1: PRIME-ANC: path-conditioned FIR synthesis from the magnitude base $A_{\mathrm{base}}=|W_{\mathrm{reg}}|$ and a shared neural generator.

Figure 2: Frequency-resolved NR (mean

\pm 1

split SD). Gray: 40-s weighted FxLMS per path/noise (PANDAR: original measurements).

Figure 3: Common GN updates (a,b) and PANDAR core design time (c). Shading:

\pm 1

split SD.

Method	Ten-path 3/7	PANDAR 7/16
PRIME-ANC, no interpolation	$16.72\!\pm\!2.08$	$16.22\!\pm\!0.66$
PRIME-ANC, paired interpolation	$\mathbf{18.81}\!\pm\!2.57$	$\mathbf{17.76}\!\pm\!0.30$
Ratio-base FIR (zero residual)	$15.95\!\pm\!0.29$	$10.09\!\pm\!0.22$
Projected complex-ratio FIR	$5.95\!\pm\!0.63$	$7.09\!\pm\!0.28$
FOMAML, pre-adaptation	$9.43\!\pm\!0.65$	$9.35\!\pm\!0.22$
Support-shared WMMSE	$7.07\!\pm\!2.46$	$8.81\!\pm\!0.27$

Table 1: No query updates: NR (dB), mean

\pm

SD over ten splits.

Output construction controls
Filter construction	NR	AMP
PRIME-ANC (standard)	17.755	0.134
Ratio-base skip, MP	18.577	0.093
Ratio-base skip, MP + phase	18.616	0.093
No skip, MP	14.646	0.093
No skip, MP + phase	14.655	0.093

Table 2: PANDAR construction and conditioning: NR/AMP (dB).

Method	NR	AMP	RMS	Time (ms)
PRIME, direct	17.755	0.134	0.302	0.717
PRIME + GN(1)	20.869	0.136	0.303	1.317
PRIME + GN(2)	21.370	0.149	0.306	1.772
PRIME + GN(3)	21.430	0.156	0.307	2.169
Ratio-base + GN(4)	19.286	0.412	0.310	2.079
WMMSE, CPU direct	21.722	0.351	0.502	4.623

Table 3: PANDAR quality and core design time (NR/AMP in dB). WMMSE + GN starts from the CPU direct solution.

Adaptive Control

Adaptive control lets systems update models, policies, or controller parameters online to maintain performance as dynamics and disturbances change.

See all 148 papers

Speech-preserving active noise control: a deep learning approach in reverberant environments
Apr 13, 2026Shuning Dai
SCHOOL OF ELECTRICAL AND ELECTRONIC ENGINEERING
Active noise cancellation on open-ear smart glasses
Apr 7, 2026Kuang Yuan, Freddy Yifei Liu, Tong Xiao +5Microphone Array Processing
Department of Electrical and Computer Engineering, Carnegie Mellon University, Pittsburgh, PA, USA · Department of Medical Physics and Acoustics, Carl von Ossietzky Universität Oldenburg, Germany · College of Computer Science and Technology, Zhejiang University, Hangzhou, China
End-to-End Multi-Task Learning for Adjustable Joint Noise Reduction and Hearing Loss Compensation
Mar 20, 2026Philippe Gonzalez, Vera Margrethe Frederiksen, Torsten Dau +1Speech Enhancement Multi-Task Learning
The authors are with the Department of Health Technology, Technical University of Denmark, 2800 Lyngby, Denmark.

Apr 13, 2026J/K move · Enter open · S save

Speech-preserving active noise control: a deep learning approach in reverberant environments

Shuning Dai

SCHOOL OF ELECTRICAL AND ELECTRONIC ENGINEERING

Open paper AskPDF arXiv

Traditional Active Noise Control (ANC) systems are mostly based on FxLMS algorithms, but such algorithms rely on linear assumptions and are often limited in handling broadband non-stationary noise or nonlinear acoustic paths. Not only that, the traditional method is used to eliminating all signals together, and noise reduction often accidentally damages the voice signal and affects normal communication. To tackle these issues, this study proposes a speech preserving deep learning ANC system, which aims to achieve stable noise reduction while effectively retaining speech in a complex acoustic environment. This study builds an end-to-end control architecture, the core of which adopts a Convolutional Recurrent Network (CRN). The structure uses the long short-term memory (LSTM) network to capture the time-related characteristics of acoustic signals. Combined with complex spectrum mapping (CSM) technology, the nonlinear distortion problem is effectively solved. In order to retain useful voice while removing noise, this study also designs a special voice retention loss function. This design guidance model selectively retains the target voice while suppressing environmental noise by identifying the characteristics of the spectrum structure. In addition, in order to verify whether the system is effective in real scenes, we use the Image Source Method (ISM) to build a high-fidelity acoustic simulation environment, which also simulates the real reverberation effect. Experimental results demonstrate that the proposed Deep ANC system achieves significantly better noise reduction than the traditional FxLMS algorithm, especially for non-stationary noises like crowd babble. Meanwhile, PESQ and STOI based evaluations confirm that the system preserves both the naturalness and intelligibility of the target speech.

Abstract

Adaptive Control

Explore similar work

Speech-preserving active noise control: a deep learning approach in reverberant environments

Active noise cancellation on open-ear smart glasses

End-to-End Multi-Task Learning for Adjustable Joint Noise Reduction and Hearing Loss Compensation