AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

DOGR: Towards Versatile Visual Document Grounding and Referring

Yinan Zhou, Yuxin Chen, Haokun Lin, Yichen Wu, Shuyu Yang, Zhongang Qi, Chen Ma, Li Zhu

ICCV 2025 · Vision-language and multimodal › Multimodal LLMs · approach: Benchmarks and analysis of MLLMs · 14 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add multi-granular parsing data in pre-training (ablation)DOGR · DocLocal4K word localizationIoU@0.5 accuracy: 74.79 -> 81.85Improved
add multi-granular parsing data in pre-training (ablation)DOGR · DocLocal4K line recognitionBLEU-4: 73.88 -> 77.60Improved
add multi-granular parsing data in pre-training (ablation)DOGR · DocLocal4K paragraph localizationIoU@0.5 accuracy: 95.46 -> 95.67No clear effect
add multi-granular parsing data in pre-training (ablation)DOGR · DOGR-Local15k chart groundingaccuracy: 24.03 -> 86.12Improved
add multi-granular parsing data in pre-training (ablation)DOGR · DOGR-Local15k poster groundingaccuracy: 63.19 -> 93.53Improved
add DOGR instruction-tuning data to DocOwl-1.5 data (ablation)DOGR · InfoVQAANLS: 64.58 -> 67.45Improved
add DOGR instruction-tuning data to DocOwl-1.5 data (ablation)DOGR · VisualMRCCIDEr: 265.9 -> 287.25Improved
add DOGR instruction-tuning data to DocOwl-1.5 data (ablation)DOGR · DocVQAANLS: 87.57 -> 89.24Improved
use accurate re-rendered bboxes instead of perturbed bboxes in chart data (ablation)DOGR · DOGR-Bench chart GrF1_all: 20.2 -> 32.91Improved
generate chart QA with GPT-4o instead of Gemini 1.5 flash (ablation)DOGR · DOGR-Bench chart GrF1_all: 1.10 -> 32.91Improved
add 25k GPT-4o chart data with perturbed bboxes (ablation)DOGR · ChartQAaccuracy: 80.92 -> 81.42Improved
add 25k GPT-4o chart data with accurate bboxes (ablation)DOGR · ChartQAaccuracy: 80.92 -> 81.88Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy