AI Paper Atlas

Research map / Detection and segmentation / CNN features

Learning to Match Anchors for Visual Object Detection

Xiaosong Zhang, Fang Wan, Chang Liu, Xiangyang Ji, Qixiang Ye

TPAMI 2021 · Detection and segmentation › CNN features · approach: Region-based and query-based detectors · 508 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
replace IoU-based assignment with FreeAnchor MLE matchingRetinaNet ResNet-50 · COCOAP: 35.7 -> 38.7Improved
replace IoU-based assignment with FreeAnchor MLE matchingRetinaNet ResNet-101 · COCOAP: 37.8 -> 40.9Improved
add FreeAnchor learning-to-matchRetinaNet ResNet-50 · COCO valNMS recall (NR): 81.8 -> 83.8Improved
add FreeAnchor learning-to-matchRetinaNet ResNet-50 · COCO valNR90: 51.3 -> 53.1Improved
add FreeAnchor learning-to-matchRetinaNet · COCOtraining time (h): 5.02h -> 5.27hNo clear effect
apply FreeAnchor to slender objectsRetinaNet · COCO slender objectsAPImproved
apply FreeAnchor to square objectsRetinaNet · COCO square objectsAPNo clear effect
increase number of objects per image (crowdedness)FreeAnchor vs RetinaNet · COCOAP gapImproved
set anchor bag size to 50 among {40, 50, 60, 100} (ablation)FreeAnchor · COCOAP: bag size 50 bestImproved
set background IoU threshold to 0.6 among {0.5, 0.6, 0.7} (ablation)FreeAnchor · COCOAP: 0.6 bestImproved
extend scale range from [640, 800] to [480, 960] (ablation)FreeAnchor ResNeXt-101 · COCO test-devAP: 44.9 -> 46.0Improved
add multi-scale testing over six scales (ablation)FreeAnchor* ResNeXt-101 · COCO test-devAP: 46.0 -> 47.3Improved

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