AI Paper Atlas

Research map / Detection and segmentation / CNN features

InfraNet: Quality-Aware RGB Guidance for Infrared Object Detection

Zichao Feng, Haodong Zhu, JingYing Yang, Linlin Yang, Yangyang Ren, Sheng Xu, Yuguang Yang, Xuhui Liu, Juan Zhang, Tian Wang, Baochang Zhang

ECCV 2026 · Detection and segmentation › CNN features · approach: RGB-D salient object detection

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 IR amplification (alpha(q)) (ablation)QualGate · LLVIPmAP: 67.2 -> 68.8Improved
add RGB suppression (w_rgb) (ablation)QualGate · LLVIPmAP: 68.4 -> 68.8Improved
add learned reliability score q instead of q=1 (ablation)InfraNet-IR · LLVIPmAP: 67.7 -> 68.8Improved
use QualGate instead of CAGF (ablation)InfraNet-IR · LLVIPmAP: 66.8 -> 68.8Improved
use QualGate instead of CRLM (ablation)InfraNet-IR · LLVIPmAP: 66.7 -> 68.8Improved
use QualGate instead of HalluciDet pseudo-RGB guidance (ablation)InfraNet-IR · LLVIPmAP: 64.9 -> 68.8Improved
add learned reliability control under fog degradation level 1.0 (ablation)InfraNet · LLVIP RGB fogAP50/AP: 92.6/61.8 -> 97.1/67.8Improved
add learned reliability control under brightness degradation level 1.0 (ablation)InfraNet · LLVIP RGB brightnessAP50/AP: 93.6/63.9 -> 96.5/66.4Improved
add auxiliary RGB supervision at training for IR-only deploymentYOLOv8 IR baseline · M3FDAP50: 84.8 -> 86.2Improved
add RGB input at inference (RGB-IR)InfraNet-IR · M3FDAP50: 86.2 -> 89.9Improved
auxiliary weight w_aux 0.25 versus 0.0 (ablation)InfraNet-IR · LLVIPmAP: 68.0 -> 68.8Improved
raise auxiliary weight w_aux from 0.25 to 1.0 (ablation)InfraNet-IR · LLVIPmAP: 68.8 -> 68.1Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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