Pith. sign in

REVIEW 25 cited by

ReFocus: Visual Editing as a Chain of Thought for Structured Image Understanding

Not yet reviewed by Pith; the record is open.

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 2501.05452 v1 pith:MGYOWXK3 submitted 2025-01-09 cs.CV cs.CL

ReFocus: Visual Editing as a Chain of Thought for Structured Image Understanding

classification cs.CV cs.CL
keywords visualimagerefocustaskseditingllmsmultimodalstructured
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

Structured image understanding, such as interpreting tables and charts, requires strategically refocusing across various structures and texts within an image, forming a reasoning sequence to arrive at the final answer. However, current multimodal large language models (LLMs) lack this multihop selective attention capability. In this work, we introduce ReFocus, a simple yet effective framework that equips multimodal LLMs with the ability to generate "visual thoughts" by performing visual editing on the input image through code, shifting and refining their visual focuses. Specifically, ReFocus enables multimodal LLMs to generate Python codes to call tools and modify the input image, sequentially drawing boxes, highlighting sections, and masking out areas, thereby enhancing the visual reasoning process. We experiment upon a wide range of structured image understanding tasks involving tables and charts. ReFocus largely improves performance on all tasks over GPT-4o without visual editing, yielding an average gain of 11.0% on table tasks and 6.8% on chart tasks. We present an in-depth analysis of the effects of different visual edits, and reasons why ReFocus can improve the performance without introducing additional information. Further, we collect a 14k training set using ReFocus, and prove that such visual chain-of-thought with intermediate information offers a better supervision than standard VQA data, reaching a 8.0% average gain over the same model trained with QA pairs and 2.6% over CoT.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 25 Pith papers

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

  1. DeepLatent: Think with Images via Parallel Latent Visual Reasoning

    cs.CV 2026-05 unverdicted novelty 7.0

    DeepLatent introduces a parallel latent visual reasoning framework with learnable 2D tokens and continuous RL, trained via distillation then RL, plus a new 180K dataset, claiming SOTA benchmark results.

  2. STORM: Internalized Modeling for Spatial-Temporal Reasoning in Video-Language Models

    cs.CV 2026-05 unverdicted novelty 7.0

    STORM teaches LVLMs to internalize spatial-temporal reasoning via bounded latent trajectories trained with generated thought videos in two stages, improving accuracy on VideoMME, MVBench and similar benchmarks while l...

  3. TableVision: A Large-Scale Benchmark for Spatially Grounded Reasoning over Complex Hierarchical Tables

    cs.AI 2026-04 conditional novelty 7.0

    TableVision benchmark shows explicit spatial grounding recovers MLLM reasoning on hierarchical tables, delivering 12.3% accuracy improvement through a decoupled perception-reasoning framework.

  4. Reasoning Within the Mind: Dynamic Multimodal Interleaving in Latent Space

    cs.CV 2025-12 unverdicted novelty 7.0

    DMLR performs dynamic visual-textual interleaving in latent space using confidence-guided latent policy gradient optimization and a dynamic visual injection strategy, yielding improved multimodal reasoning on benchmarks.

  5. Training Multi-Image Vision Agents via End2End Reinforcement Learning

    cs.CV 2025-12 unverdicted novelty 7.0

    IMAgent trains a multi-image vision agent via pure end-to-end RL with visual reflection tools and a two-layer motion trajectory masking strategy, reaching SOTA on single- and multi-image benchmarks while revealing too...

  6. Latent Visual Reasoning

    cs.CV 2025-09 unverdicted novelty 7.0

    Latent Visual Reasoning enables autoregressive generation of latent visual states that reconstruct critical image tokens, yielding gains on perception-heavy VQA benchmarks such as 71.67% on MMVP.

  7. DeFacto: Counterfactual Thinking with Images for Enforcing Evidence-Grounded and Faithful Reasoning

    cs.AI 2025-09 unverdicted novelty 7.0

    DeFacto trains multimodal models using counterfactual image variants and reinforcement learning rewards to improve both answer accuracy and evidence-answer consistency.

  8. Visual-TableQA: Open-Domain Benchmark for Reasoning over Table Images

    cs.CV 2025-09 conditional novelty 7.0

    Visual-TableQA is a new open-domain benchmark of rendered table images and complex QA pairs created via multi-LLM collaborative generation, with fine-tuned models showing robust generalization to external tests.

  9. VAD: Attributing Visual Evidence for Target Reconstruction in Multimodal On-Policy Distillation

    cs.CV 2026-07 conditional novelty 6.0

    Counterfactual present/removed teacher views attribute visually supported corrections and reconstruct student-anchored distillation targets that beat source-mixed multimodal OPD.

  10. FaithEyes: Towards Faithful Tool Use via Multi-Agent Process-Image Verification

    cs.CV 2026-07 conditional novelty 6.0

    A self-judging multi-agent VLM injects process-image helpfulness verdicts into tool observations and scales tool rewards by the helpful-call ratio, improving accuracy and tool faithfulness.

  11. Dive into the Scene: Breaking the Perceptual Bottleneck in Vision-Language Decision Making via Focus Plan Generation

    cs.CV 2026-06 unverdicted novelty 6.0

    SceneDiver introduces a coarse-to-fine focus plan generation approach for VLMs that constructs holistic scene graphs then iteratively decomposes tasks, plus a distillation adapter for VLAs, to reduce visual hallucinat...

  12. Vision-OPD: Learning to See Fine Details for Multimodal LLMs via On-Policy Self-Distillation

    cs.CV 2026-05 unverdicted novelty 6.0

    Vision-OPD transfers an MLLM's privileged regional perception to its full-image policy through on-policy token-level self-distillation, yielding competitive results on fine-grained visual benchmarks.

  13. Vision-OPD: Learning to See Fine Details for Multimodal LLMs via On-Policy Self-Distillation

    cs.CV 2026-05 unverdicted novelty 6.0

    Vision-OPD uses on-policy self-distillation from crop-conditioned to full-image policies within the same MLLM to close the regional-to-global perception gap.

  14. Meta-CoT: Enhancing Granularity and Generalization in Image Editing

    cs.CV 2026-04 unverdicted novelty 6.0

    Meta-CoT uses two-level decomposition of editing operations into meta-tasks and a CoT consistency reward to improve granularity and generalization, reporting 15.8% gains across 21 tasks.

  15. Fully Spiking Neural Networks with Target Awareness for Energy-Efficient UAV Tracking

    cs.CV 2026-03 conditional novelty 6.0

    A two-stage pure RL method with an information-gap global view and hierarchical grounding loss makes MLLMs truly rely on precise crops and sets SOTA on high-res VQA under tight token budgets.

  16. DeFacto: Counterfactual Thinking with Images for Enforcing Evidence-Grounded and Faithful Reasoning

    cs.AI 2025-09 unverdicted novelty 6.0

    DeFacto trains multimodal models with counterfactual image variants and GRPO reinforcement learning to enforce that correct answers are supported by correct visual evidence.

  17. Grounded Reinforcement Learning for Visual Reasoning

    cs.CV 2025-05 unverdicted novelty 6.0

    ViGoRL introduces visually grounded RL that anchors reasoning steps to image coordinates and uses multi-turn zooming to outperform standard RL and supervised baselines on spatial and GUI reasoning benchmarks.

  18. From Structure to Synergy: A Survey of Vision-Language Perception Paradigm Evolution in Multimodal Large Language Models

    cs.CL 2026-06 unverdicted novelty 5.0

    The survey formalizes MLLM perception as a unified vision-language capability and traces its evolution via a new five-stage taxonomy while outlining future challenges.

  19. V-Zero: Answer-Label-Free On-Policy Distillation with Contrastive Evidence Gating for Fine-Grained Visual Reasoning

    cs.CV 2026-06 unverdicted novelty 5.0

    V-Zero trains MLLMs for visual reasoning without answer labels by gating on-policy distillation trajectories using contrastive evidence from relevant versus negative image crops.

  20. Latent Visual States for Efficient Multimodal Reasoning

    cs.CV 2026-06 unverdicted novelty 5.0

    EVA generates adaptive Latent_slot tokens as internal visual thoughts, trained end-to-end with text tokens via D-GSPO on the EVA-230K dataset, claiming performance gains and better inference efficiency.

  21. MedLVR: Latent Visual Reasoning for Reliable Medical Visual Question Answering

    cs.CV 2026-04 unverdicted novelty 5.0

    MedLVR interleaves latent visual reasoning segments in autoregressive decoding and uses two-stage training to raise average medical VQA accuracy from 48.3% to 53.4% over a Qwen2.5-VL-7B backbone on OmniMedVQA and five...

  22. MedLVR: Latent Visual Reasoning for Reliable Medical Visual Question Answering

    cs.CV 2026-04 conditional novelty 5.0

    Latent visual reasoning with ROI-supervised fine-tuning and latent-space policy optimization raises average medical VQA accuracy from 48.3% to 53.4% on five external benchmarks.

  23. Decompose, Look, and Reason: Reinforced Latent Reasoning for VLMs

    cs.CL 2026-04 unverdicted novelty 5.0

    DLR is a new reinforced latent reasoning method for VLMs that decomposes queries, uses continuous visual latents, and outperforms text-only and multimodal CoT baselines on vision-centric benchmarks with better interpr...

  24. Learning to Focus and Precise Cropping: A Reinforcement Learning Framework with Information Gaps and Grounding Loss for MLLMs

    cs.CV 2026-03 unverdicted novelty 5.0

    A two-stage RL method with information gaps and grounding loss trains MLLMs to focus on and precisely crop relevant image regions, yielding SOTA results on high-resolution VQA benchmarks.

  25. Multimodal Chain-of-Thought Reasoning: A Comprehensive Survey

    cs.CV 2025-03 unverdicted novelty 2.0

    The paper provides the first comprehensive survey of multimodal chain-of-thought reasoning, including foundational concepts, a taxonomy of methodologies, application analyses, challenges, and future directions.