AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

Customized Fusion: A Closed-Loop Dynamic Network for Adaptive Multi-Task-Aware Infrared-Visible Image Fusion

Zengyi Yang, Yu Liu, Juan Cheng, Zhiqin Zhu, Yafei Zhang, Huafeng Li

CVPR 2026 · Vision-language and multimodal › Cross-modal learning · approach: Multi-modality image fusion · 2 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
apply semantic compensation (RSC) to the frozen fusion network (ablation)CLDyN · M3FD object detection (YOLOv5)mAP50-95: 0.6076 -> 0.6304Improved
apply RSC compensation to a different detector without retraining (ablation)CLDyN · M3FD detection (DETR)mAP50-95: 0.5610 -> 0.5810Improved
add closed-loop optimization mechanism (vs Model I) (ablation)CLDyN · M3FD object detectionmAP50-95: 0.6272 -> 0.6304Improved
add closed-loop optimization mechanism (vs Model I) (ablation)CLDyN · VT5000 SODmFbeta: 0.8136 -> 0.8129No clear effect
add reward loss (vs Model II) (ablation)CLDyN · M3FD object detectionmAP50-95: 0.6276 -> 0.6304Improved
replace plain conv networks with RSC (vs Model III) (ablation)CLDyN · FMB segmentationmIoU: 60.07 -> 60.34Improved
use CLDyN instead of joint-training fusion methodsCLDyN · multi-task (OD, Seg, SOD)parameters: 0.46M vs 39.95M-134.97MImproved
use CLDyN vs MRFS joint training on its trained taskCLDyN · FMB segmentationmIoU: 61.48 -> 60.34Hurt
use CLDyN fusion quality vs prior methodsCLDyN · M3FD fusionQAB/F: 0.6601 -> 0.6900Improved
use CLDyN fusion on VT5000CLDyN · VT5000 fusionMI: 2.5749 -> 2.3352Hurt

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