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

Maxime Burchi, Valentin Vielzeuf