Efficient Conformer: Progressive Downsampling and Grouped Attention for Automatic Speech Recognition.

Maxime Burchi, Valentin Vielzeuf

Browse the full ASRU paper archive.