AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

Spatially Consistent Representation Learning

Byungseok Roh, Wuhyun Shin, Ildoo Kim, Sungwoong Kim

CVPR 2021 · Detection and segmentation › Open-set and few-shot detection · approach: Object-level representation learning · 79 citations (OpenAlex)

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 spatially matched local representations instead of global image-level matchingSelf-supervised pretraining transferred to object detection · COCO val2017, Faster R-CNN and RetinaNetAP: 40.0 -> 40.9 (Faster R-CNN); 37.7 -> 39.0 (RetinaNet)Improved
Use spatially matched local representations instead of global image-level matchingSelf-supervised pretraining transferred to localization tasks · PASCAL VOC test2007 detectionAP: 55.0 -> 57.2Improved
Increase the number of matched RoIs from 1 to 10 (ablation)SCRL · COCO val2017 detectionAP: 39.2 -> 40.2Improved
Add exact spatial matching of the boxes rather than jittering them (ablation)SCRL with 1 or 10 RoIs · COCO val2017 detectionAP: 39.2 -> 39.1 (10% jitter); 39.2 -> 38.8 (20% jitter); 40.2 -> 40.0 (10% jitter); 40.2 -> 39.6 (20% jitter)Improved
Add IoU thresholding to reject redundant sampled boxes (ablation)SCRL with 10 RoIs · COCO val2017 detectionAP: 39.8 -> 40.2Improved
Set pretraining learning rate to 0.3 instead of 0.45 (ablation)SCRL · COCO val2017 detectionAP: 40.9 -> 41.2; +0.3 APImproved
Add negative pairs using the SimCLR framework (ablation)SCRL upstream training · COCO detectionAP: +0.47 APImproved
Use input-level spatial consistency instead of feature-level matching (ablation)SCRL-style pretraining · Localization downstream tasksAPHurt
Pretrain with SCRL rather than BYOLResNet-50 · COCO detection with 10% training dataAP: 25.5 -> 26.4Improved
Pretrain with SCRL rather than BYOLResNet-50 · COCO val2017 detection schedules from 0.5× to 7×AP: 37.0 -> 38.3; 40.0 -> 40.9; 41.6 -> 42.3; 41.7 -> 42.5; 41.6 -> 42.6Improved

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