# Spatial-Magnifier: Spatial upsampling for multichannel speech enhancement Source: [Spatial-Magnifier: Spatial upsampling for multichannel speech enhancement]() Collection index: [llms.txt]() Data updated: 2026-06-24T05:00:04.550Z Authors: Dongheon Lee, Ashutosh Pandey, Sanjeel Parekh, Daniel Wong, Jacob Donley, Buye Xu, Juan Azcarreta Published: 2026-05-06T10:55:30.000Z ## Abstract While the spatial directivity of multichannel speech enhancement algorithms improves with the number of microphones, fitting large capture arrays into real-world edge devices is typically limited by physical constraints. To overcome this limitation, we propose Spatial-Magnifier, a neural network designed to generate virtual microphone (VM) signals from a limited set of real microphone (RM) measurements. Moreover, we introduce the Spatial Audio Representation Learning (SARL) framework, which leverages estimated VM signals and features to condition a downstream speech enhancement system. Experimental results demonstrate that the proposed framework outperforms existing spatial upsampling baselines across various speech extraction systems, including end-to-end multichannel speech enhancement and neural beamforming. The proposed method nearly recovers the oracle performance achieved when all microphones are available. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Omnidirectional Speech Enhancement on Omni-SE RM: 2ch, VM: 1ch](): Omnidirectional Speech Enhancement; Omni-SE RM: 2ch, VM: 1ch; SI-SDR (VM-BF): 6.87. - [Omnidirectional Speech Enhancement on Omni-SE RM: 2ch, VM: 4ch](): Omnidirectional Speech Enhancement; Omni-SE RM: 2ch, VM: 4ch; SI-SDR (VM-BF): 8.37. ## Original sources and implementation [Original paper]() [PDF]()