AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

InfoScan: Information-Efficient Visual Scanning via Resource-Adaptive Walks

Yifeng Wu, Huimin Huang, Shangjie Zhou, Yawen Huang, Hao Zheng, Yun Chen, Xian Wu, Ruize Han, Guanhua Chen

ICLR 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Attention-based visual token pruning

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 InfoScan, an information-aware, resource-adaptive scanning mechanism for state-space (Mamba-style) visual backbones to replace fixed scan patterns in high-resolution visual representation learning.

Problem

High-resolution visual representation learning is costly due to quadratic ViT complexity, and fixed scanning patterns in Mamba-based vision models hinder content-adaptive perception.

Approach

An information-aware scanning mechanism for state-space visual backbones that dynamically adapts scan walks to image content and resource budget, instead of using fixed scan orders.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add patch selection module (ablation)InfoScan · ImageNet-1K 512x512Top-1 (%): 82.5 -> 83.4Improved
add path planning module on top of patch selection (ablation)InfoScan · ImageNet-1K 512x512Top-1 (%): 83.4 -> 85.9Improved
add adaptive revisitation reward (ablation)InfoScan scanning policy · ImageNet-1KTop-1 (%): 80.4 -> 81.1Improved
add exploration bonus to reward (ablation)InfoScan scanning policy · ImageNet-1KTop-1 (%): 81.1 -> 84.2Improved
add neighborhood information gain to reward (ablation)InfoScan scanning policy · ImageNet-1KTop-1 (%): 84.2 -> 85.9Improved
add Shannon entropy to information scoring (ablation)Information Scoring Module · ImageNet-1KTop-1 (%): 79.7 -> 81.6Improved
add local variance to entropy scoring (ablation)Information Scoring Module · ImageNet-1KTop-1 (%): 81.6 -> 85.9Improved
repeat the fixed raster scan three times (ablation)Raster scan · ImageNet-1K and ADE-20KTop-1 / mIoU: Top-1 83.5 -> 83.6; ADE-20K 42.9 -> 41.8No clear effect
repeat the fixed Hilbert scan three times (ablation)Hilbert scan · ImageNet-1K, ADE-20K, BraTSTop-1 / mIoU: Top-1 83.6 -> 84.8; BraTS 17.1 -> 16.8Mixed
randomize the scan start point (ablation)Single Hilbert scan · ImageNet-1KTop-1 (%): 83.6 -> 84.5Improved
randomize the scan start point (ablation)InfoScan · ImageNet-1KTop-1 (%): 84.2 -> 85.9Improved
replace fixed scan with learned InfoScan policy (fixed start) (ablation)Triple Hilbert Scan · ImageNet-1K and ADE-20KTop-1 / mIoU: Top-1 84.8 vs 84.2; ADE-20K 44.2 vs 44.6Mixed

Experimental setup

Builds on: Mamba Vision Mamba VMamba

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy