AI Paper Atlas

Research map / Detection and segmentation / CNN features

Towards Balanced Learning for Instance Recognition

Jiangmiao Pang, Kai Chen, Qi Li, Zhihai Xu, Huajun Feng, Jianping Shi, Wanli Ouyang, Dahua Lin

IJCV 2021 · Detection and segmentation › CNN features · approach: Region-based and query-based detectors · 27 citations (OpenAlex)

Read the paper

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 IoU-balanced sampling (ablation)Faster R-CNN R50-FPN · COCO val-2017AP: 35.9 -> 36.8Improved
add balanced feature pyramid (ablation)Faster R-CNN with IoU-balanced sampling · COCO val-2017AP: 36.8 -> 37.7Improved
add LGR-L1 loss (ablation)Libra R-CNN · COCO val-2017AP: 37.7 -> 38.5Improved
add LGR-CE loss (ablation)Libra R-CNN · COCO val-2017AP: 38.5 -> 39.0Improved
vary number of IoU bins K (2, 3, 5) (ablation)IoU-balanced sampling · COCO val-2017AP: 36.7 / 36.8 / 36.7No clear effect
balance positive samples per ground truth (ablation)Faster R-CNN · COCO val-2017AP: 35.9 -> 36.1No clear effect
add non-local refinement to balanced semantic features (ablation)Balanced feature pyramid · COCO val-2017AP: 36.3 -> 36.8Improved
add balanced semantic features to PAFPN (ablation)PAFPN · COCO val-2017AP: 36.3 -> 37.2Improved
raise localization loss weight from 1.5 to 2.0 (ablation)Faster R-CNN · COCO val-2017AP: 36.4 -> 36.2Hurt
replace balanced L1 (alpha 0.5, gamma 1.0) with gamma 1.5 (ablation)Balanced L1 loss · COCO val-2017AP: 36.5 -> 37.2Improved
use plain L1 loss instead of balanced L1 (ablation)Faster R-CNN · COCO val-2017AP: 36.6 vs 37.2Hurt
add bucketed class re-weighting (BCR) (ablation)Libra R-CNN · LVIS valAPr: 14.5 -> 16.8Improved

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