4 citations · 7 across the 6 of their papers we have counts for
1 paper · 1 filter
Peter R. D. van der Wal, Nicola Strisciuglio, George Azzopardi
Vision Transformers (ViTs) have demonstrated remarkable performance in computer vision tasks. However, their self-attention mechanism often diffuses focus across background regions…