NeuralKalman: A Learnable Kalman Filter for Acoustic Echo Cancellation

Yixuan Zhang,Meng Yu,Hao Zhang,Dong Yu,DeLiang Wang

2023-12-26

Abstract:The robustness of the Kalman filter to double talk and its rapid convergence make it a popular approach for addressing acoustic echo cancellation (AEC) challenges. However, the inability to model nonlinearity and the need to tune control parameters cast limitations on such adaptive filtering algorithms. In this paper, we integrate the frequency domain Kalman filter (FDKF) and deep neural networks (DNNs) into a hybrid method, called NeuralKalman, to leverage the advantages of deep learning and adaptive filtering algorithms. Specifically, we employ a DNN to estimate nonlinearly distorted far-end signals, a transition factor, and the nonlinear transition function in the state equation of the FDKF algorithm. Experimental results show that the proposed NeuralKalman improves the performance of FDKF significantly and outperforms strong baseline methods.

Audio and Speech Processing,Sound

What problem does this paper attempt to address?

The paper aims to address the challenging issues in Acoustic Echo Cancellation (AEC). Specifically, the goal of the paper is to develop a hybrid model by introducing Deep Neural Networks (DNN) based on the traditional Kalman filtering algorithm, to fully leverage the advantages of the frequency domain Kalman filtering algorithm and DNN. The research focuses on the following aspects: 1. **Nonlinear Distortion**: Traditional adaptive filtering algorithms fail to effectively handle the nonlinear distortion caused by speakers. 2. **Covariance Estimation**: In Kalman filtering, the covariance of state noise and observation noise is usually approximated, which may affect the performance of the algorithm. 3. **Transition Factor**: In the state equation of Kalman filtering, the transition factor is usually assigned a fixed value, which makes it difficult to adapt to environmental changes. To address the above issues, the paper proposes the NeuralKalman framework, which utilizes DNN to estimate nonlinear distortion, transition factors, and nonlinear transformation functions in the state equation. Experimental results show that the proposed NeuralKalman method has better robustness in double-talk scenarios and outperforms existing baseline methods in echo suppression.

NeuralKalman: A Learnable Kalman Filter for Acoustic Echo Cancellation

Low-Complexity Acoustic Echo Cancellation with Neural Kalman Filtering

Dynamic Neural Network-based Solutions for Acoustic Echo Suppression

Neural Network Augmented Kalman Filter for Robust Acoustic Howling Suppression

NEURAL KALMAN FILTERING FOR SPEECH ENHANCEMENT

Acoustic Echo Cancellation Algorithm Based on Kalman Filtering of Skewed Observation Noise

Joint Acoustic Echo Cancellation and Speech Dereverberation Using Kalman filters

End-To-End Deep Learning-based Adaptation Control for Linear Acoustic Echo Cancellation

Deep Neural Network Based Regression Approach for Acoustic Echo Cancellation

Nonlinear Acoustic Echo Cancellation with Deep Learning

Reinforcement Learning Adaptive Kalman Filter for AE Signal’s AR-mode Denoise

Acoustic Echo Cancellation by Combining Adaptive Digital Filter and Recurrent Neural Network

Adaptive KalmanNet: Data-Driven Kalman Filter with Fast Adaptation

NeuralEcho: A Self-Attentive Recurrent Neural Network For Unified Acoustic Echo Suppression And Speech Enhancement

A Robust and Cascaded Acoustic Echo Cancellation Based on Deep Learning

Neuralecho: Hybrid of Full-Band and Sub-Band Recurrent Neural Network For Acoustic Echo Cancellation and Speech Enhancement.

Hybrid AHS: A Hybrid of Kalman Filter and Deep Learning for Acoustic Howling Suppression

AEC in a NetShell: On Target and Topology Choices for FCRN Acoustic Echo Cancellation

Joint Neural AEC and Beamforming with Double-Talk Detection

Time-Variance Aware Dynamic Kernel Generation for Real-Time Acoustic Echo Cancellation

A Two-stage Progressive Neural Network for Acoustic Echo Cancellation.