AI Paper Atlas

Research map / Vision-language and multimodal / Visual question answering

Learning to Focus and Precise Cropping:A Reinforcement Learning Framework with Information Gaps and Grounding Loss for MLLMs

Xuanpu Zhao, Zhentao Tan, Dianmo Sheng, Tianxiang Chen, Yao Liu, Yue Wu, Tao Gong, Qi Chu, Nenghai Yu

CVPR 2026 · Vision-language and multimodal › Visual question answering · approach: Multimodal LLM question answering · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add information gap (downsampled global image, full-res crop) in stage I (ablation)Qwen2.5-VL-7B RL agent · HR-Bench 8K (1,024 tokens)overall accuracy: 61.0 -> 69.9Improved
add information gap in stage I (ablation)Qwen2.5-VL-7B RL agent · V* (1,024 tokens)overall accuracy: 72.3 -> 78.0Improved
add grounding reward stage II with VisualProbe data and L1 reward (ablation)Qwen2.5-VL-7B RL agent · HR-Bench 8K (1,024 tokens)overall accuracy: 69.9 -> 72.1Improved
stage II trained on first-stage source data instead of VisualProbe (ablation)Qwen2.5-VL-7B RL agent · HR-Bench 8K (1,024 tokens)overall accuracy: 69.9 -> 68.1Hurt
add L1 distance reward on top of VisualProbe data (ablation)Qwen2.5-VL-7B RL agent · V* (1,024 tokens)overall accuracy: 76.4 -> 80.6Improved
add L1 distance reward on top of VisualProbe data (ablation)Qwen2.5-VL-7B RL agent · HR-Bench 8K (1,024 tokens)overall accuracy: 72.7 -> 72.1No clear effect
replace crop with GT or noise to test crop reliance (information gap training) (ablation)Qwen2.5-VL-7B RL agent · HR-Bench 8Kaccuracy delta: 4.5 -> 18.5Improved
use 1,024-token input vs 16,384 for competitors, time per questionQwen2.5-VL-7B RL agent · HR-Bench 8Kseconds per question: 12.4 (DeepEyes) -> 2.8Improved
replace DeepEyes crop with ground truth box or noise (preliminary)Qwen2.5-VL-7B RL agent · HR-Bench 8K (16,384 tokens)overall accuracy: 68.0 vs 69.5 GT vs 68.5 noiseNo clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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