AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

Self-Supervised Representation Learning via Neighborhood-Relational Encoding

Mohammad Sabokrou, Mohammad Khalooei, Ehsan Adeli

ICCV 2019 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Invariance-based self-supervised learning · 39 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 neighborhood-relational loss to an auto-encoderauto-encoder, linear SVM · MNISTaccuracy: 0.969 -> 0.978Improved
add neighborhood-relational loss to an auto-encoderauto-encoder, RBF SVM · MNISTaccuracy: 0.972 -> 0.984Improved
use NRE-pretrained AlexNet instead of AE pretrainingAlexNet self-supervised · PASCAL VOC 2007 classification (all layers)mAP: 53.8 -> 74.4Improved
use NRE-pretrained AlexNet instead of DeepClusterAlexNet self-supervised · PASCAL VOC 2007 classification (all layers)mAP: 73.7 -> 74.4Improved
use NRE-pretrained AlexNet instead of DeepClusterAlexNet self-supervised · PASCAL VOC 2007 detectionmAP: 55.4 -> 54.7No clear effect
use NRE-pretrained AlexNet instead of DeepClusterAlexNet self-supervised · PASCAL VOC 2012 segmentationmIoU: 45.1 -> 51.1Improved
use NRE-pretrained AlexNet instead of supervised AlexNetAlexNet · PASCAL VOC 2012 segmentationmIoU: 48.0 -> 51.1Improved
use NRE-pretrained AlexNet instead of supervised AlexNetAlexNet · PASCAL VOC 2007 classification (all layers)mAP: 78.3 -> 74.4Improved
replace MagNet reconstruction with NRE reconstruction as FGSM defenseMNIST classifier · MNIST FGSM epsilon=0.3accuracy: 0.4242 -> 0.7166Improved
replace MagNet with NRE as FGSM defenseMNIST classifier · MNIST FGSM epsilon=0.01accuracy: 0.9238 -> 0.9802Improved
drop the lambda3 far-sample term (0.6,0.4,0) versus (0.6,0.2,0.2) (ablation)NRE · UCSD Ped2EER (%): 17.5% -> 14%Improved
use NRE single auto-encoder instead of Deep-cascadeanomaly detector · UCSD Ped2EER (%): 9% -> 14%Hurt

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