AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Open Set Learning with Counterfactual Images

Lawrence Neal, Matthew Olson, Xiaoli Fern, Weng-Keen Wong, Fuxin Li

ECCV 2018 · Label-efficient and robust learning › Recognition and long-tail · approach: Loss design and classifier theory · 512 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 counterfactual open set images as an extra classsoftmax classifier · SVHN open set detection (K=6, M=10)ROC AUC: .886 -> .910Improved
add counterfactual open set images as an extra classsoftmax classifier · CIFAR-10 open set detection (K=6, M=10)ROC AUC: .677 -> .699Improved
add counterfactual open set images as an extra classsoftmax classifier · MNIST open set detection (K=6, M=10)ROC AUC: .978 -> .988Improved
use counterfactual generation instead of OpenMaxclassifier · SVHN open set detection (K=6, M=10)ROC AUC: .894 -> .910Improved
use counterfactual optimization instead of latent interpolation (G-OpenMax*)synthetic open set generation · CIFAR-10 open set detection (K=6, M=10)ROC AUC: .675 -> .699Improved
use counterfactual optimization instead of latent interpolation (G-OpenMax*)synthetic open set generation · Tiny-ImageNet extended open setROC AUC: .580 -> .586Improved
use counterfactual optimization instead of latent interpolation (G-OpenMax*)synthetic open set generation · CIFAR+50 extended open setROC AUC: .819 -> .827Improved
add counterfactual images to training, effect on known-class accuracyclassifier · CIFAR-10 closed set accuracy (K=6)accuracy: .801 -> .821Improved
add counterfactual images to training, effect on known-class accuracyclassifier · MNIST closed set accuracy (K=6)accuracy: .995 -> .996No clear effect
use generative OpenMax (interpolation) rather than OpenMaxOpenMax · CIFAR-10 open set detection (K=6, M=10)ROC AUC: .695 -> .675No clear effect
use G-OpenMax* instead of OpenMax on digit datasetsOpenMax · MNIST open set detection (K=6, M=10)ROC AUC: .981 -> .984Improved
use G-OpenMax* instead of OpenMaxOpenMax · CIFAR+50 extended open setROC AUC: .796 -> .819Improved

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