# Insights Into Deep Non-linear Filters for Improved Multi-channel Speech Enhancement arXiv: https://arxiv.org/abs/2206.13310 ยท HTML: https://mlsanity.com/paper/2206.13310/ ## Key facts - Cited by 14 papers in the map (data through 14 Oct 2026). - Builds on 17 of its 53 references that are in the map. - First posted on arXiv 27 Jun 2022. ## Builds on (most cited first, 17 of 17) - [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014) - [Pyroomacoustics: A Python package for audio room simulations and array processing algorithms](https://mlsanity.com/paper/1710.04196/) (2017) - [The fifth 'CHiME' Speech Separation and Recognition Challenge: Dataset, task and baselines](https://mlsanity.com/paper/1803.10609/) (2018) - [FullSubNet: A Full-Band and Sub-Band Fusion Model for Real-Time Single-Channel Speech Enhancement](https://mlsanity.com/paper/2010.15508/) (2020) - [End-to-end Microphone Permutation and Number Invariant Multi-channel Speech Separation](https://mlsanity.com/paper/1910.14104/) (2019) - [FaSNet: Low-latency Adaptive Beamforming for Multi-microphone Audio Processing](https://mlsanity.com/paper/1909.13387/) (2019) - [ADL-MVDR: All deep learning MVDR beamformer for target speech separation](https://mlsanity.com/paper/2008.06994/) (2020) - [Channel-Attention Dense U-Net for Multichannel Speech Enhancement](https://mlsanity.com/paper/2001.11542/) (2020) - [DF-Conformer: Integrated architecture of Conv-TasNet and Conformer using linear complexity self-attention for speech enhancement](https://mlsanity.com/paper/2106.15813/) (2021) - [Embedding and Beamforming: All-neural Causal Beamformer for Multichannel Speech Enhancement](https://mlsanity.com/paper/2109.00265/) (2021) - [Narrow-band Deep Filtering for Multichannel Speech Enhancement](https://mlsanity.com/paper/1911.10791/) (2019) - [Nonlinear Spatial Filtering in Multichannel Speech Enhancement](https://mlsanity.com/paper/2104.11033/) (2021) - [TPARN: Triple-path Attentive Recurrent Network for Time-domain Multichannel Speech Enhancement](https://mlsanity.com/paper/2110.10757/) (2021) - [On the Role of Spatial, Spectral, and Temporal Processing for DNN-based Non-linear Multi-channel Speech Enhancement](https://mlsanity.com/paper/2206.11181/) (2022) - [Complex-valued Spatial Autoencoders for Multichannel Speech Enhancement](https://mlsanity.com/paper/2108.03130/) (2021) - [Multichannel Speech Enhancement by Raw Waveform-mapping using Fully Convolutional Networks](https://mlsanity.com/paper/1909.11909/) (2019) - [Inter-channel Conv-TasNet for multichannel speech enhancement](https://mlsanity.com/paper/2111.04312/) (2021) ## Built on by (most cited first, 14 of 14) - [Multi-channel Speech Separation Using Spatially Selective Deep Non-linear Filters](https://mlsanity.com/paper/2304.12023/) (2023) - [ReZero: Region-customizable Sound Extraction](https://mlsanity.com/paper/2308.16892/) (2023) - [McNet: Fuse Multiple Cues for Multichannel Speech Enhancement](https://mlsanity.com/paper/2211.08872/) (2022) - [Spatially Selective Deep Non-linear Filters for Speaker Extraction](https://mlsanity.com/paper/2211.02420/) (2022) - [Multi-Microphone Speaker Separation by Spatial Regions](https://mlsanity.com/paper/2303.07143/) (2023) - [Distilling HuBERT with LSTMs via Decoupled Knowledge Distillation](https://mlsanity.com/paper/2309.09920/) (2023) - [Speech enhancement with frequency domain auto-regressive modeling](https://mlsanity.com/paper/2309.13537/) (2023) - [LMFCA-Net: A Lightweight Model for Multi-Channel Speech Enhancement with Efficient Narrow-Band and Cross-Band Attention](https://mlsanity.com/paper/2502.11462/) (2025) - [LipDiffuser: Lip-to-Speech Generation with Conditional Diffusion Models](https://mlsanity.com/paper/2505.11391/) (2025) - [Self-Steering Deep Non-Linear Spatially Selective Filters for Efficient Extraction of Moving Speakers under Weak Guidance](https://mlsanity.com/paper/2507.02791/) (2025) - [Leveraging Joint Spectral and Spatial Learning with MAMBA for Multichannel Speech Enhancement](https://mlsanity.com/paper/2409.10376/) (2024) - [Autoregressive Guidance of Deep Spatially Selective Filters using Bayesian Tracking for Efficient Extraction of Moving Speakers](https://mlsanity.com/paper/2603.23723/) (2026) - [End-to-End Multi-Microphone Speaker Extraction Using Relative Transfer Functions](https://mlsanity.com/paper/2502.06285/) (2025) - [Comparison of Knowledge Distillation Methods for Low-complexity Multi-microphone Speech Enhancement using the FT-JNF Architecture](https://mlsanity.com/paper/2507.19208/) (2025) ## Abstract (arXiv) The key advantage of using multiple microphones for speech enhancement is that spatial filtering can be used to complement the tempo-spectral processing. In a traditional setting, linear spatial filtering (beamforming) and single-channel post-filtering are commonly performed separately. In contrast, there is a trend towards employing deep neural networks (DNNs) to learn a joint spatial and tempo-spectral non-linear filter, which means that the restriction of a linear processing model and that of a separate processing of spatial and tempo-spectral information can potentially be overcome. However, the internal mechanisms that lead to good performance of such data-driven filters for multi-channel speech enhancement are not well understood. Therefore, in this work, we analyse the properties of a non-linear spatial filter realized by a DNN as well as its interdependency with temporal and spectral processing by carefully controlling the information sources (spatial, spectral, and temporal) available to the network. We confirm the superiority of a non-linear spatial processing model, which outperforms an oracle linear spatial filter in a challenging speaker extraction scenario for a low number of microphones by 0.24 POLQA score. Our analyses reveal that in particular spectral information should be processed jointly with spatial information as this increases the spatial selectivity of the filter. Our systematic evaluation then leads to a simple network architecture, that outperforms state-of-the-art network architectures on a speaker extraction task by 0.22 POLQA score and by 0.32 POLQA score on the CHiME3 data.