Direction-Preserving MIMO Speech Enhancement Using a Neural Covariance Estimator

Multichannel speech enhancement is widely used as a front-end in microphone array processing systems. While most existing approaches produce a single enhanced signal, direction-preserving multiple-input multiple-output (MIMO) methods instead aim to provide enhanced multichannel signals that retain directional properties, enabling downstream applications such as beamforming, binaural rendering, and direction-of-arrival estimation. In this work, we propose a fully blind, direction-preserving MIMO speech enhancement method based on neural estimation of the spatial noise covariance matrix. A lightweight OnlineSpatialNet estimates a scale-normalized Cholesky factor of the frequency-domain noise covariance, which is combined with a direction-preserving MIMO Wiener filter to enhance speech while preserving the spatial characteristics of both target and residual noise. In contrast to prior approaches relying on oracle information or mask-based covariance estimation for single-output systems, the proposed method directly targets accurate multichannel covariance estimation with low computational complexity. Experimental results show improved speech enhancement, covariance estimation capability, and performance in downstream tasks over a mask-based baseline, approaching oracle performance with significantly fewer parameters and computational cost.

Residual Learning for Neural Ambisonics Encoders

Emerging wearable devices such as smartglasses and extended reality headsets demand high-quality spatial audio capture from compact, head-worn microphone arrays. Ambisonics provides a device-agnostic spatial audio representation by mapping array signals to spherical harmonic (SH) coefficients. In practice, however, accurate encoding remains challenging. While traditional linear encoders are signal-independent and robust, they amplify low-frequency noise and suffer from high-frequency spatial aliasing. On the other hand, neural network approaches can outperform linear encoders but they often assume idealized microphones and may perform inconsistently in real-world scenarios. To leverage their complementary strengths, we introduce a residual-learning framework that refines a linear encoder with corrections from a neural network. Using measured array transfer functions from smartglasses, we compare a UNet-based encoder from the literature with a new recurrent attention model. Our analysis reveals that both neural encoders only consistently outperform the linear baseline when integrated within the residual learning framework. In the residual configuration, both neural models achieve consistent and significant improvements across all tested metrics for in-domain data and moderate gains for out-of-domain data. Yet, coherence analysis indicates that all neural encoder configurations continue to struggle with directionally accurate high-frequency encoding.

Blind Identification of Binaural Room Impulse Responses from Smart Glasses

Smart glasses are increasingly recognized as a key medium for augmented reality, offering a hands-free platform with integrated microphones and non-ear-occluding loudspeakers to seamlessly mix virtual sound sources into the real-world acoustic scene. To convincingly integrate virtual sound sources, the room acoustic rendering of the virtual sources must match the real-world acoustics. Information about a user's acoustic environment however is typically not available. This work uses a microphone array in a pair of smart glasses to blindly identify binaural room impulse responses (BRIRs) from a few seconds of speech in the real-world environment. The proposed method uses dereverberation and beamforming to generate a pseudo reference signal that is used by a multichannel Wiener filter to estimate room impulse responses which are then converted to BRIRs. The multichannel room impulse responses can be used to estimate room acoustic parameters which is shown to outperform baseline algorithms in the estimation of reverberation time and direct-to-reverberant energy ratio. Results from a listening experiment further indicate that the estimated BRIRs often reproduce the real-world room acoustics perceptually more convincingly than measured BRIRs from other rooms with similar geometry.

Direct and Residual Subspace Decomposition of Spatial Room Impulse Responses

Psychoacoustic experiments have shown that directional properties of, in particular, the direct sound, salient reflections, and the late reverberation of an acoustic room response can have a distinct influence on the auditory perception of a given room. Spatial room impulse responses (SRIRs) capture those properties and thus are used for direction-dependent room acoustic analysis and virtual acoustic rendering. This work proposes a subspace method that decomposes SRIRs into a direct part, which comprises the direct sound and the salient reflections, and a residual, to facilitate enhanced analysis and rendering methods by providing individual access to these components. The proposed method is based on the generalized singular value decomposition and interprets the residual as noise that is to be separated from the other components of the reverberation. It utilizes a noise estimate to identify large generalized singular values, which are then attributed to the direct part. By advancing from the end of the SRIR toward the beginning while iteratively updating the noise estimate, the method is able to work with anisotropic and slowly time-varying reverberant sound fields. The proposed method does not require direction-of-arrival estimation of reflections and shows an improved separation of the direct part from the residual compared to an existing approach. A case study with measured SRIRs suggests a high robustness of the method under different acoustic conditions. A reference implementation is provided.