Reproduce “A Task-Centric Survey of Long-Document Visual Question Answering: Taxonomy, Evolution, and Future Directions” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yufei Tong, Guanjie Cheng, Tianle Wu, Leilei Zhang, Hang Zheng, Peihan Wu, Jianwei Yin, Shuiguang Deng
The authors survey long-document visual question answering, organizing over 150 papers around parsing, evidence retrieval, generation, and evaluation. They distinguish explicit retrieval, which selects material for generation, from direct reading of supplied representations. About 80% of surveyed methods use explicit retrieval. This is a classification of the literature, not evidence that retrieval outperforms direct reading.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.