Pith. sign in

REVIEW 9 cited by

ZoomEye: Enhancing Multimodal LLMs with Human-Like Zooming Capabilities through Tree-Based Image Exploration

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 2411.16044 v4 pith:LDE6LJ2D submitted 2024-11-25 cs.CV

Haozhan Shen , Kangjia Zhao , Tiancheng Zhao , Ruochen Xu , Zilun Zhang , Mingwei Zhu , Jianwei Yin This is my paper · ORCID
classification cs.CV
keywords mllmsreasoningvisualimagemodelszoomlargealgorithm
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Multimodal Large Language Models (MLLMs) have demonstrated impressive capabilities in vision-language understanding. Recently, with the integration of test-time scaling techniques, these models have also shown strong potential in visual reasoning. However, most existing reasoning approaches remain text-level in nature: MLLMs are prompted to explore various combinations of textual tokens via their underlying language model, while the visual input remains fixed throughout the reasoning process. This paradigm limits the model's ability to fully exploit rich visual information, particularly when dealing with images containing numerous fine-grained elements. In such cases, vision-level reasoning becomes crucial - where models dynamically zoom into specific regions of the image to gather detailed visual cues necessary for accurate decision-making. In this paper, we propose Zoom Eye, a training-free, model-agnostic tree search algorithm tailored for vision-level reasoning. Zoom Eye treats an image as a hierarchical tree structure, where each child node represents a zoomed-in sub-region of its parent, and the root corresponds to the full image. The algorithm enables MLLMs to simulate human-like zooming behavior by navigating from root to leaf nodes in search of task-relevant visual evidence. We experiment on a series of high-resolution benchmarks and the results demonstrate that Zoom Eye consistently improves the performance of multiple MLLMs by a large margin (e.g., InternVL2.5-8B increases by 15.71% and 17.69% on HR-Bench) and also enables small 3-8B MLLMs to outperform strong large models such as GPT-4o. Code: https://github.com/om-ai-lab/ZoomEye

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. HiDe: Rethinking The Zoom-IN method in High Resolution MLLMs via Hierarchical Decoupling

    cs.CV 2025-09 unverdicted novelty 7.0 of 10

    HiDe uses token-wise attention decoupling and layout-preserving decoupling to build compact crops that push Qwen2.5-VL and InternVL3 to state-of-the-art scores on high-resolution VQA benchmarks.

  2. R-VLM: Region-Aware Vision Language Model for Precise GUI Grounding

    cs.CV 2025-07 conditional novelty 7.0 of 10

    R-VLM improves GUI grounding by combining two-stage zoom-in proposals with an IoU-weighted training loss, raising accuracy by up to 13 absolute points over SeeClick.

  3. AdaTurn: Budget-Aware Test-Time Scaling for Active Visual Perception Agents

    cs.CV 2026-07 conditional novelty 6.0 of 10

    Budget-conditioned forced-answer RL lifts 4-turn VisualProbe-Medium from 36.7% to 47.6% while keeping 32-turn performance competitive.

  4. WeaveEarth: Structured Evidence Construction and Reasoning for Training-Free UHR Remote Sensing Understanding

    cs.CV 2026-07 conditional novelty 6.0 of 10

    A training-free evidence-selection and topology-preserving packaging pipeline improves frozen VLMs on ultra-high-resolution remote sensing VQA without multi-round search.

  5. BVS: Bayesian Visual Search with Multimodal Large Language Model for Fine-grained Perception

    cs.CV 2026-07 conditional novelty 6.0 of 10

    BVS combines early-stop attention rollout priors with a scale-aware non-stationary kernel and GP-UCB to locate tiny objects in UHR images more accurately and with fewer MLLM queries than prior visual-search methods.

  6. Zoom-Refine: Boosting High-Resolution Multimodal Understanding via Localized Zoom and Self-Refinement

    cs.CV 2025-06 conditional novelty 6.0 of 10

    A training-free two-step pipeline, localize a bounding box then refine the answer from a high-resolution crop, improves MLLM accuracy on high-resolution image benchmarks.

  7. Iterative Zoom-In: Temporal Interval Exploration for Long Video Understanding

    cs.CV 2025-06 conditional novelty 5.0 of 10

    A confidence-guided iterative zoom-in framework improves long-video question answering for multimodal LLMs without any training.

  8. Reinforcing VLMs to Use Tools for Detailed Visual Reasoning Under Resource Constraints

    cs.LG 2025-06 conditional novelty 5.0 of 10

    A 3B VLM trained with GRPO to call a zoom tool improves V*Bench accuracy by 5.7% over its base model but degrades TextVQA and HR-Bench performance.

  9. Unifying Language Agent Algorithms with Graph-based Orchestration Engine for Reproducible Agent Research

    cs.CL 2025-05 conditional novelty 5.0 of 10

    AGORA is a graph-based agent framework that standardizes ten reasoning algorithms; its evaluations show simple Chain-of-Thought prompting is often the most cost-effective, though without statistical rigor.

Pith tools