AI Paper Atlas

Research map / Detection and segmentation / CNN features

Detecting Backdoors in Object Detection via Pre-NMS Prediction Distribution Shift

Longtian Wang, Chenhao Lin, Zhengyu Zhao, Le Yang, Shiwei Wang, Yuhan Zhi, Xiaofei Xie, Chao Shen

ECCV 2026 · Detection and segmentation › CNN features · approach: Region-based and query-based detectors · 0 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
Use equal-size single-object crops as validation set (ablation)DistScan (Faster R-CNN) · Val-set ablation, VOC, Faster R-CNNavg detection accuracy (%): 67.59 -> 96.30Improved
Use equal-size crops instead of full images (ablation)DistScan (YOLOv5) · Val-set ablation, VOC, YOLOv5avg detection accuracy (%): 100.00 -> 88.89Hurt
Enforce equal per-class counts vs random images (ablation)DistScan (YOLOv5) · Val-set ablation, VOC, YOLOv5avg detection accuracy (%): 100.00 -> 100.00No clear effect
Apply confidence threshold 0.0005 vs none (ablation)DistScan (Faster R-CNN) · Confidence threshold ablation, VOC, Faster R-CNNdetection accuracy (%): 50.00 -> 96.30Improved
Raise confidence threshold above 0.001 (ablation)DistScan (Faster R-CNN) · Confidence threshold ablation, COCO, Faster R-CNNdetection accuracy (%): 95.37 -> 67.59Hurt
Raise confidence threshold to 0.01 for YOLOv5 on COCO (ablation)DistScan (YOLOv5) · Confidence threshold ablation, COCO, YOLOv5detection accuracy (%): 96.30 -> 97.22No clear effect
DistScan vs ODSCAN on scene-level attacksDistScan · Scene-level attacks (SPONGE, BLINDING, GMA)detection accuracy (%): 50.00 -> 96.99 avgImproved
DistScan vs MIA on GMA COCO Faster R-CNNDistScan · GMA attack, COCO, Faster R-CNNdetection accuracy (%): 58.33 -> 100.00Improved
DistScan on additional attacks/architecturesDistScan · VOC extra attacks (YOLOv8, DETR, BadDet)mean AUROC: 0.73 -> 0.87Improved
JS divergence vs KL, L2, cosine distance (ablation)DistScan · Distance metric (appendix)detection performanceImproved

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