Pith. sign in

REVIEW 9 cited by

ViDoRAG: Visual Document Retrieval-Augmented Generation via Dynamic Iterative Reasoning Agents

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2502.18017 v2 pith:L6RYV4JJ submitted 2025-02-25 cs.CV cs.AIcs.CLcs.IR

Qiuchen Wang , Ruixue Ding , Zehui Chen , Weiqi Wu , Shihang Wang , Pengjun Xie , Feng Zhao This is my paper · ORCID
classification cs.CVcs.AIcs.CLcs.IR
keywords reasoningvidoragvisualdocumentsmethodsretrievalvidoseekapproaches
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Understanding information from visually rich documents remains a significant challenge for traditional Retrieval-Augmented Generation (RAG) methods. Existing benchmarks predominantly focus on image-based question answering (QA), overlooking the fundamental challenges of efficient retrieval, comprehension, and reasoning within dense visual documents. To bridge this gap, we introduce ViDoSeek, a novel dataset designed to evaluate RAG performance on visually rich documents requiring complex reasoning. Based on it, we identify key limitations in current RAG approaches: (i) purely visual retrieval methods struggle to effectively integrate both textual and visual features, and (ii) previous approaches often allocate insufficient reasoning tokens, limiting their effectiveness. To address these challenges, we propose ViDoRAG, a novel multi-agent RAG framework tailored for complex reasoning across visual documents. ViDoRAG employs a Gaussian Mixture Model (GMM)-based hybrid strategy to effectively handle multi-modal retrieval. To further elicit the model's reasoning capabilities, we introduce an iterative agent workflow incorporating exploration, summarization, and reflection, providing a framework for investigating test-time scaling in RAG domains. Extensive experiments on ViDoSeek validate the effectiveness and generalization of our approach. Notably, ViDoRAG outperforms existing methods by over 10% on the competitive ViDoSeek benchmark. The code is available at https://github.com/Alibaba-NLP/ViDoRAG.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 9 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. HiEviDR-Bench: A Benchmark for Hierarchical Evidence Aggregation in Deep Research

    cs.IR 2026-07 conditional novelty 6.5 of 10

    A hierarchical evidence-graph benchmark reveals that multimodal deep-research models write fluent reports while failing citation, claim, and answer grounding.

  2. M$^3$Prune: Hierarchical Collaborative Pruning for Efficient Multi-Modal Multi-Agent Retrieval-Augmented Generation

    cs.MM 2026-08 conditional novelty 6.0 of 10

    M3Prune learns hierarchical communication graphs across text and image agents, prunes redundant links, and reports higher accuracy at lower token cost on three multimodal QA benchmarks.

  3. SlideAgent: Hierarchical Agentic Framework for Multi-Page Visual Document Understanding

    cs.CL 2025-10 unverdicted novelty 6.0 of 10

    SARA combines natural-language snippets with semantic compression vectors in RAG to improve answer relevance, correctness, and similarity on 9 datasets across 5 LLMs.

  4. TextlessRAG: End-to-End Visual Document RAG by Speech Without Text

    cs.CV 2025-09 conditional novelty 6.0 of 10

    TextlessRAG retrieves document image pages and generates spoken answers directly from an audio query, without OCR, ASR, or TTS, and introduces a bilingual speech-document RAG benchmark.

  5. MM-BrowseComp: A Comprehensive Benchmark for Multimodal Browsing Agents

    cs.CL 2025-08 unverdicted novelty 6.0 of 10

    MM-BrowseComp is a multimodal browsing benchmark on which the best evaluated model, OpenAI o3 with tools, achieves 29.02% accuracy.

  6. MaskSearch: A Universal Pre-Training Framework to Enhance Agentic Search Capability

    cs.CL 2025-05 conditional novelty 6.0 of 10

    A pre-training task called RAMP, where models practice searching to fill masked text spans, improves downstream agentic open-domain QA performance across Qwen and LLaMA models.

  7. Magic-MM-Embedding: Towards Visual-Token-Efficient Universal Multimodal Embedding with MLLMs

    cs.CV 2026-02 conditional novelty 5.0 of 10

    Visual token compression (4x fewer tokens) plus a three-stage generative/contrastive/judge-curated training pipeline yields state-of-the-art MLLM-based retrieval accuracy at lower inference cost.

  8. VisRAG2.0: Mitigating Visual Hallucinations via Evidence-Guided Multi-Image Reasoning in Visual Retrieval-Augmented Generation

    cs.CL 2025-10 conditional novelty 5.0 of 10

    EVisRAG, an evidence-guided observe-record-reason-answer pipeline trained with reward-scoped GRPO, improves multi-image visual QA accuracy by about 19% over its backbone VLM across five benchmarks.

  9. VRAG-RL: Empower Vision-Perception-Based RAG for Visually Rich Information Understanding via Iterative Reasoning with Reinforcement Learning

    cs.CL 2025-05 conditional novelty 5.0 of 10

    VRAG-RL uses GRPO reinforcement learning with visual cropping actions and a retrieval-aware reward to improve vision-language RAG agents on document benchmarks.

Pith tools