A saliency-guided warp-unwarp method reallocates spatial representation to preserve fine structures in latent diffusion models for image-to-image translation.
Constructive Distortion: Improving MLLMs with Attention-Guided Image Warping
2 Pith papers cite this work. Polarity classification is still indexing.
abstract
Multimodal large language models (MLLMs) often miss small details and spatial relations in cluttered scenes, leading to errors in fine-grained perceptual grounding. We introduce AttWarp, a lightweight method that allocates more resolution to query-relevant content while compressing less informative areas, all while preserving global context. At test time, the approach uses an MLLM's cross-modal attention to perform rectilinear warping of the input image, reallocating spatial resolution toward regions the model deems important, without changing model weights or architecture. This attention-guided warping preserves all original image information but redistributes it non-uniformly, so small objects and subtle relationships become easier for the same model to read while the global layout remains intact. Across five benchmarks (TextVQA, GQA, DocVQA, POPE, MMMU) and four MLLMs (LLaVA, Qwen-VL, InternVL, and InstructBLIP), AttWarp consistently improves accuracy, strengthens compositional reasoning, and reduces hallucinations, outperforming four competitive baselines that manipulate raw images at test time. Together, these results show that attention-guided warping prioritizes information relevant to the query while preserving context, and that the same MLLMs perform better when given such warped inputs.
fields
cs.CV 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
OPPO is an evidence-aware preference optimization objective that contrasts faithful responses under varying visual evidence strengths to reduce hallucinations in MLLMs.
citing papers explorer
-
WarpI2I: Image Warping for Image-to-Image Translation
A saliency-guided warp-unwarp method reallocates spatial representation to preserve fine structures in latent diffusion models for image-to-image translation.
-
Clearer Sight, Fewer Lies: Oriented Pickup Preference Optimization for Multimodal Hallucination Mitigation
OPPO is an evidence-aware preference optimization objective that contrasts faithful responses under varying visual evidence strengths to reduce hallucinations in MLLMs.