Transformer-based Neural Beamforming for Real-Time Speech Enhancement on Smart Low-Power Hearable Devices
This work presents an optimized methodology for the real-time execution of a neural-network-based minimum variance distortionless response (MVDR) beamformer on MCUs, and is believed to be the first real-time multi-channel Transformer-based neural beamforming pipeline deployed on an MCU-class device.
Luca Bompani, M. Fariselli, Giovanni Oltrecolli et al.
· 0 citations