Abstract: Document Visual Question Answering (DocVQA) offers a promising approach to extracting insights from large document corpora. However, existing benchmarks focus on evaluating multi-modal ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results