English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Recent document question answering models consist of two key components: thevision encoder, which captures layout and visual elements in images, and aLarge Language Model (LLM) that helps contextualize questions to the image andsupplements them with external world knowledge to generate accurate answers.However, the relative contributions of the vision encoder and the languagemodel in these tasks remain unclear. This is especially interesting given theeffectiveness of instruction-tuned LLMs, which exhibit remarkable adaptabilityto new tasks. To this end, we explore the following aspects in this work: (1)The efficacy of an LLM-only approach on document question answering tasks (2)strategies for serializing textual information within document images andfeeding it directly to an instruction-tuned LLM, thus bypassing the need for anexplicit vision encoder (3) thorough quantitative analysis on the feasibilityof such an approach. Our comprehensive analysis encompasses six diversebenchmark datasets, utilizing LLMs of varying scales. Our findings reveal thata strategy exclusively reliant on the LLM yields results that are on par withor closely approach state-of-the-art performance across a range of datasets. Weposit that this evaluation framework will serve as a guiding resource forselecting appropriate datasets for future research endeavors that emphasize thefundamental importance of layout and image content information.

Analyzing the Efficacy of an LLM-Only Approach for Image-based Document Question Answering