Attention Where It Matters: Rethinking Visual Document Understanding with Selective Region Concentration

Open in new window