AI Paper Atlas

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

Object-aware Contrastive Learning for Debiased Scene Representation

Sangwoo Mo, Hyunwoo Kang, Kihyuk Sohn, Chun-Liang Li, Jinwoo Shin

NeurIPS 2021 · Detection and segmentation › Open-set and few-shot detection · approach: Object-level representation learning · 16 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
add object-aware random crop using ContraCAM boxesMoCov2 (ResNet-50, COCO-trained) · CIFAR100 linear evalaccuracy (%): 53.26 -> 62.72Improved
add object-aware random crop using ContraCAM boxesBYOL (ResNet-50) · CIFAR10 linear evalaccuracy (%): 76.62 -> 82.79Improved
add OA-Crop with ContraCAM boxesMoCov2 (ResNet-50) · COCO detectionAP (%): 36.34 -> 36.60No clear effect
use ground-truth boxes instead of ContraCAM boxes for OA-Crop (ablation)MoCov2 (ResNet-50) · Food linear evalaccuracy (%): 62.01 -> 57.47Hurt
use ground-truth boxes instead of ContraCAM boxes for OA-Crop (ablation)BYOL (ResNet-50) · CIFAR10 linear evalaccuracy (%): 82.79 -> 85.92Improved
add OA-Crop (CAM) when evaluating on in-distribution ImageNet-9MoCov2 (ResNet-50) · ImageNet-9accuracy (%): 84.67 -> 84.54No clear effect
add OA-Crop (CAM) for distribution shiftMoCov2 (ResNet-50) · ImageNet-Sketch-9accuracy (%): 41.44 -> 43.11Improved
add background mixup using ContraCAMMoCov2 (ResNet-18) · Background Challenge Mixed-Randaccuracy (%): 60.34 -> 66.89Improved
add background mixup using ContraCAMMoCov2 (ResNet-18) · Background Challenge BG-Gapgap (%): 20.64 -> 17.24Improved
use BG-Mixup instead of standard Mixup or CutMix (ablation)MoCov2 (ResNet-18) · ImageNet-C-9accuracy (%): 40.15 -> 44.39Improved
use BG-Mixup instead of Mixup (ablation)BYOL (ResNet-18) · Stylized-ImageNet-9accuracy (%): 28.11 -> 27.01Hurt
increase ContraCAM iterations from 1 to 10 (ablation)ContraCAM · COCO mask mIoUmask mIoU: 0.081 -> 0.319Improved

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