AI Paper Atlas

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

Distilling Localization for Self-Supervised Representation Learning

Nanxuan Zhao, Zhirong Wu, Rynson W.H. Lau, Stephen Lin

AAAI 2021 · Detection and segmentation › Open-set and few-shot detection · approach: Object-level representation learning · 31 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 saliency-guided copy-and-paste background augmentationInstDist, CMC, MoCo-v1, and MoCo-v2 · ImageNet validationlinear accuracy: 56.5 -> 59.3 and 62.9; 63.4 -> 65.0 and 66.9; 60.6 -> 62.8 and 65.0; 67.5 -> 67.9 and 69.2Improved
Use class-agnostic saliency masks for copy-and-paste augmentation (ablation)MoCo · ImageNet validationlinear accuracy: 60.6 -> 62.7; 60.6 -> 62.1; 60.6 -> 62.8; 60.6 -> 65.0Improved
Apply copy-and-paste augmentation to 30%, 50%, or 70% of images rather than none (ablation)MoCo · ImageNet validationlinear accuracy: 60.6 -> 62.8; 60.6 -> 62.2; 60.6 -> 61.6Improved
Apply copy-and-paste augmentation to every image (ablation)MoCo · ImageNet validationlinear accuracy: 60.6 -> 47.6Hurt
Replace backgrounds with texture images (ablation)MoCo with DiLo · ImageNet validationlinear accuracy: 60.6 -> 60.6 (+0.0)No clear effect
Replace backgrounds with ImageNet crops (ablation)MoCo with DiLo · ImageNet validationlinear accuracy: 60.6 -> 62.1Improved
Replace backgrounds with homogeneous grayscale images (ablation)MoCo with DiLo · ImageNet validationlinear accuracy: 60.6 -> 62.8Improved
Use Gaussian blending or mixed blending when pasting (ablation)MoCo with DiLo · ImageNet validationlinear accuracy: 60.6 -> 62.5; 60.6 -> 62.8Improved
Mask pooled final-layer features using saliency instead of training with copy-and-paste augmentations (ablation)MoCo · ImageNet validationlinear accuracy: -19%Hurt
Add intense color jittering (ablation)MoCo self-supervised representation · ImageNet validationlinear accuracy: 40.4 -> 56.9Improved
Add intense color jittering (ablation)Supervised representation · ImageNet validationlinear accuracy: 77.5 -> 77.4Hurt
Add RBD or BASNet background augmentationMoCo · VOC 2007 testAP: 55.9 -> 56.5; 55.9 -> 56.9Improved

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