Through the Magnifying Glass: Adaptive Perception Magnification for Hallucination-Free VLM Decoding

[2503.10183] Through the Magnifying Glass: Adaptive Perception Magnification for Hallucination-Free VLM Decoding Skip to main content Search Submit Donate Log in Search arXiv Press Enter to search · Advanced search Computer Science > Computer Vision and Pattern Recognition arXiv:2503.10183 (cs) [Submitted on 13 Mar 2025 ( v1 ), last revised 9 Apr 2026 (this version, v4)] Title: Through the Magnifying Glass: Adaptive Perception Magnification for Hallucination-Free VLM Decoding Authors: Shunqi Mao , Chaoyi Zhang , Weidong Cai View a PDF of the paper titled Through the Magnifying Glass: Adaptive Perception Magnification for Hallucination-Free VLM Decoding, by Shunqi Mao and 2 other authors View PDF HTML (experimental) Abstract: Existing vision-language models (VLMs) often suffer from visual hallucination, where the generated responses contain inaccuracies that are not grounded in the visual input. Efforts to address this issue without model finetuning primarily mitigate hallucination by contrastively reducing language biases or amplifying the weights of visual embedding during decoding. However, these approaches remain limited in their ability to capture fine-grained visual details. In this work, we propose the Perception Magnifier (PM), a novel visual decoding method that iteratively isolates relevant visual tokens based on attention and magnifies the corresponding regions, spurring the model to concentrate on fine-grained visual details during decoding. By magnifying critical regions while preserving the structural and contextual information at each decoding step, PM allows the VLM to enhance its scrutiny of the visual input, hence producing more accurate and faithful responses. Extensive experimental results demonstrate that PM not only achieves superior hallucination mitigation but also enhances language generation while preserving strong reasoning capabilities. Code can be found at this https URL . Comments: ACL 2026 Main Conference Subjects: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI) Cite as: arXiv:2503.10183 [cs.CV] (or arXiv:2503.10183v4 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2503.10183 Focus to learn more arXiv-issued DOI via DataCite Submission history From: Shunqi Mao [ view email ] [v1] Thu, 13 Mar 2025 09:14:11 UTC (31,187 KB) [v2] Fri, 14 Mar 2025 01:48:33 UTC (31,187 KB) [v3] Wed, 6 Aug 2025 13:45:16 UTC (17,773 KB) [v4] Thu, 9 Apr 2026 08:49:13 UTC (15,421 KB) Full-text links: Access Paper: View a PDF of the paper titled Through the Magnifying Glass: Adaptive Perception Magnification for Hallucination-Free VLM Decoding, by Shunqi Mao and 2 other authors View PDF HTML (experimental) TeX Source view license Current browse context: cs.CV < prev | next > new | recent | 2025-03 Change to browse by: cs cs.AI References & Citations NASA ADS Google Scholar Semantic Scholar export BibTeX citation Loading... BibTeX formatted citation × loading... Data provided by: Bookmark Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer ( What is the Explorer? ) Connected Papers Toggle Connected Papers ( What is Connected Papers? ) Litmaps Toggle Litmaps ( What is Litmaps? ) scite.ai Toggle scite Smart Citations ( What are Smart Citations? ) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv ( What is alphaXiv? ) Links to Code Toggle CatalyzeX Code Finder for Papers ( What is CatalyzeX? ) DagsHub Toggle DagsHub ( What is DagsHub? ) GotitPub Toggle Gotit.pub ( What is GotitPub? ) Huggingface Toggle Hugging Face ( What is Huggingface? ) ScienceCast Toggle ScienceCast ( What is ScienceCast? ) Demos Demos Replicate Toggle Replicate ( What is Replicate? ) Spaces Toggle Hugging Face Spaces ( What is Spaces? ) Spaces Toggle TXYZ.AI ( What is TXYZ.AI? ) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower ( What are Influence Flowers? ) Core recommender toggle CORE Recommender ( What is CORE? ) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs . Which authors of this paper are endorsers? | Disable MathJax ( What is MathJax? ) We gratefully acknowledge support from our major funders , member institutions , , and all contributors. About · Help · Contact · Subscribe · Copyright · Privacy · Accessibility · Operational Status (opens in new tab) Major funding support from

Related documents

• ARKit LabelMaker: A New Scale for Indoor 3D Scene Understanding#974165
• StreamTTT: Reconciling Real-Time Perception and Long-Term Memory in Streaming VLMs#615212
• StreamTTT: Reconciling Real-Time Perception and Long-Term Memory in Streaming VLMs#614245
• Unsupervised visual structuring of polished ornamental granites using self-supervised vision transformers#12038
• Unsupervised visual structuring of polished ornamental granites using self-supervised vision transformers#12037