Peeling the Onion: Hierarchical Reduction of Data Redundancy for Efficient Vision Transformer Training

Open in new window