← back to paper
arxiv: 2509.10234 · 2 revisions
Data-independent Beamforming for End-to-end Multichannel Multi-speaker ASR