AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Don't Overthink with Pixels: Efficient Reasoning for Segmentation

Song Wang, Gongfan Fang, Lingdong Kong, Xiangtai Li, Jianyun Xu, Sheng Yang, Qiang Li, Jianke Zhu, Xinchao Wang

ICML 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Chain-of-thought multimodal reasoning

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 PixelThink, a reasoning segmentation approach that uses reinforcement fine-tuning to reduce overthinking by adapting reasoning length to task complexity.

Problem

Reasoning segmentation methods fine-tune MLLMs with image-text-mask data and reinforcement fine-tuning, but they overthink and produce uniformly verbose reasoning chains regardless of task complexity.

Approach

Abstract is truncated; the stated idea is a simple approach that makes reasoning length adaptive to task complexity during reinforcement fine-tuning for segmentation, to avoid uniformly verbose chains.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add difficulty-aware token budget (alone) (ablation)PixelThink · ReasonSeg-DIFF valgIoU / tokens: gIoU 59.65 -> 62.07; tokens 86.92 -> 39.95Improved
add uncertainty-guided budget (alone) (ablation)PixelThink · ReasonSeg-DIFF valgIoU / tokens: gIoU 59.65 -> 59.54 (cIoU 50.92 -> 52.37); tokens 86.92 -> 42.41Mixed
combine difficulty and uncertainty (ablation)PixelThink · ReasonSeg-DIFF valgIoU (%): 59.65 -> 63.81Improved
use 3-level difficulty split instead of 2 or 4 (ablation)PixelThink · ReasonSeg-DIFF valtokens / gIoU: 2-level: 61.78 tok, 64.38 gIoU; 3-level: 46.98 tok, 63.81; 4-level: 76.89 tok, 62.28Mixed
constrain length of medium-difficulty samples (ablation)PixelThink · ReasonSeg-DIFF valgIoU (%): 63.81 -> 60.92Hurt
raise uncertainty weight alpha from 25 to 35 (ablation)PixelThink · ReasonSeg-DIFF valtokens / gIoU: tokens 46.98 -> 71.10; gIoU 63.81 -> 62.66Hurt
replace top-2 margin with entropy as uncertainty (ablation)PixelThink · ReasonSeg-DIFF valgIoU (%): 63.81 -> 62.43Hurt
remove reasoning entirely (no-thinking) (ablation)Seg-Zero · ReasonSeg-DIFF valgIoU (%): 63.81 -> 60.19 (RL) / 60.37 (prompt)Hurt
add prompt-level 64-token limitSeg-Zero · ReasonSeg-DIFF valgIoU (%): 61.63 -> 62.50; tokens 90.79 -> 57.62Improved
add L1-Exact rigid length rewardSeg-Zero · ReasonSeg-DIFF valgIoU (%): 61.63 -> 40.97Hurt
add soft length penalty with difficulty and uncertaintySeg-Zero · Inference latencylatency (ms): 1282.48 -> 892.93Improved
label difficulty with direct MLLM prompting rather than three-dimension decomposition (ablation)PixelThink · ReasonSeg-DIFF valgIoU (%): 63.81 -> 56.32Hurt

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