COMCAT: Towards Efficient Compression and Customization of Attention-Based Vision Models.

Published in: CoRR (2023)

Keyphrases