MVDR Beamforming for Cyclostationary Processes

2025-10-21

AI Summary

Traditional MVDR beamformers assume short-term stationarity of noise, failing to model spectral correlations inherent in periodic noise sources such as musical instruments or fans. To address this, we propose cyclic MVDR (cMVDR), the first MVDR extension formulated for cyclostationary processes. cMVDR integrates FRESH filtering with data-driven estimation of resonant frequencies, explicitly modeling coherence among frequency-shifted inputs to jointly capture spectral and spatial correlations. This formulation mitigates harmonic misalignment issues and enables effective single-channel speech enhancement. Experiments on realistic nonstationary noise scenarios demonstrate that cMVDR improves SI-SDR by up to 5 dB over conventional MVDR, while maintaining robust, high performance even under single-microphone conditions.

Technology Category

• Machine Learning: Multi-instance/Multi-view Learning
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Computer Vision: Multi-modal Vision

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Conventional acoustic beamformers assume that noise is stationary within short time frames. This assumption prevents them from exploiting correlations between frequencies in almost-periodic noise sources such as musical instruments, fans, and engines. These signals exhibit periodically varying statistics and are better modeled as cyclostationary processes. This paper introduces the cyclic MVDR (cMVDR) beamformer, an extension of the conventional MVDR that leverages both spatial and spectral correlations to improve noise reduction, particularly in low-SNR scenarios. The method builds on frequency-shifted (FRESH) filtering, where shifted versions of the input are combined to attenuate or amplify components that are coherent across frequency. To address inharmonicity, where harmonic partials deviate from exact integer multiples of the fundamental frequency, we propose a data-driven strategy that estimates resonant frequencies via periodogram analysis and computes the frequency shifts from their spacing. Analytical and experimental results demonstrate that performance improves with increasing spectral correlation. On real recordings, the cMVDR achieves up to 5 dB gain in scale-invariant signal-to-distortion ratio (SI-SDR) over the MVDR and remains effective even with a single microphone. Code is available at https://github.com/Screeen/cMVDR.

Problem

• Enhances noise reduction for cyclostationary signals like engines
• Addresses inharmonicity through data-driven frequency shift estimation
• Improves performance in low-SNR scenarios using spectral correlations

Innovation

• Extends MVDR beamformer with cyclostationary noise modeling
• Uses frequency-shifted filtering to exploit spectral correlations
• Implements data-driven frequency shift estimation for inharmonic signals