Research map / Vision-language and multimodal / Multimodal LLMs
One Patch Doesn’t Fit All: Adaptive Patching for Native-Resolution Multimodal Large Language Models
Wenzhuo Liu, Weijie Yin, Fei Zhu, Shijie Ma, Haiyang Guo, yi chen, Xiao-Hui Li, LiangXiao, ChaoFeng, Cheng-Lin Liu
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes adaptive patching for native-resolution MLLMs so that patch granularity varies with image content, preserving fine detail for documents and charts while keeping cost low for simple images.
Problem
Native-resolution MLLMs must handle images of varying resolution; uniform patching wastes compute on simple images or loses fine detail on dense documents and charts.
Approach
Adaptive patching that chooses patch sizes per image region so that high-resolution detail is preserved only where needed, rather than using one fixed patch size for all inputs.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add adaptive patch size selection (training-free PI-resize) | Qwen2.5-VL-3B · MME | score: 2135.90 -> 2210.41 | Improved |
| add adaptive patch size selection | Qwen2.5-VL-3B · OCRBench | score: 821 -> 845 | Improved |
| add adaptive patching | Ovis2.5-2B · DocVQA val | accuracy: 93.90 -> 93.22 | Hurt |
| add adaptive patching | Qwen2.5-VL-3B · LLaVABench | score: 66.30 -> 65.34 | Hurt |
| add adaptive patching across model scales (ablation) | Qwen2.5-VL 3B/7B/32B, Ovis2.5 2B/9B · MME | score: 2313.17 -> 2345.96 (7B) | Improved |
| add adaptive patching instead of resizing images to a fixed range (ablation) | Qwen2.5-VL-3B · MME | score: 2165.01 -> 2210.41 | Improved |
| use adaptive patch sizes during SFT (training-based) | SAIL-VL-2B · ChartQA test (LLaVA1.5-665K) | accuracy: 47.96 -> 54.84 | Improved |
| use PI-resize rather than bilinear or area interpolation for patch embedding conversion (ablation) | Qwen2.5-VL-3B · pixel-range benchmarks | accuracy | Improved |
| increase alpha (resolution sensitivity) in patching law (ablation) | AdaPatch · Qwen2.5-VL-3B hyperparameter sweep | accuracy | Hurt |
| use adaptive patching at high resolution | AdaPatch · inference time | latency | Mixed |
Closest papers by meaning
- Divide, Conquer and Combine: A Training-Free Framework for High-Resolution Image Perception in Multimodal Large Language Models (AAAI 2025 · similarity 0.92)
- Res-Bench: Benchmarking the Robustness of Multimodal Large Language Models to Dynamic Resolution Input (AAAI 2026 · similarity 0.89)
- Feast Your Eyes: Mixture-of-Resolution Adaptation for Multimodal Large Language Models (ICLR 2025 · similarity 0.89)
- NaViL: Rethinking Scaling Properties of Native Multimodal Large Language Models under Data Constraints (NeurIPS 2025 · similarity 0.89)
- Constructive Distortion: Improving MLLMs with Attention-Guided Image Warping (ICLR 2026 · similarity 0.89)
- CoVFT: Context-aware Visual Fine-tuning for Multimodal Large Language Models (CVPR 2026 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
