Max-AST: Combining Convolution, Local and Global Self-Attentions for Audio Event Classification.

Tony Alex, Sara Ahmed, Armin Mustafa, Muhammad Awais, Philip J. B. Jackson

Browse the full ICASSP paper archive.