Meta-Learning for Adaptive Filters with Higher-Order Frequency Dependencies

TL;DR

This paper introduces higher-order meta-adaptive filters that leverage frequency dependencies to improve acoustic echo cancellation, achieving significant performance gains with reduced complexity.

Contribution

The work presents a novel higher-order meta-adaptive filter framework that exploits frequency dependencies, outperforming baselines in acoustic echo cancellation tasks.

Findings

Multi-dB improvements over baselines

At least ten times less complex

Effective with or without speech enhancer

Abstract

Adaptive filters are applicable to many signal processing tasks including acoustic echo cancellation, beamforming, and more. Adaptive filters are typically controlled using algorithms such as least-mean squares(LMS), recursive least squares(RLS), or Kalman filter updates. Such models are often applied in the frequency domain, assume frequency independent processing, and do not exploit higher-order frequency dependencies, for simplicity. Recent work on meta-adaptive filters, however, has shown that we can control filter adaptation using neural networks without manual derivation, motivating new work to exploit such information. In this work, we present higher-order meta-adaptive filters, a key improvement to meta-adaptive filters that incorporates higher-order frequency dependencies. We demonstrate our approach on acoustic echo cancellation and develop a family of filters that yield multi-dB improvements over competitive baselines, and are at least an order-of-magnitude less complex. Moreover, we show our improvements hold with or without a downstream speech enhancer.