Efficient conformer: Progressive downsampling and grouped attention for automatic speech recognition

Maxime Burchi, Valentin Vielzeuf