AI Paper Atlas

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

Addressing Model Vulnerability to Distributional Shifts Over Image Transformation Sets

Riccardo Volpi, Vittorio Murino

ICCV 2019 · Label-efficient and robust learning › Recognition and long-tail · approach: Correspondence and causal representation learning · 112 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 worst-case transformation search (RS) augmentation on top of random augmentation (ablation)RDA MNIST ConvNet · MNIST RS-found transformationsaccuracy: .941 -> .977Improved
add worst-case transformation search (ES) augmentation on top of random augmentation (ablation)RDA MNIST ConvNet · MNIST ES-found transformationsaccuracy: .927 -> .977Improved
add random transformation augmentation (RDA) to ERM (ablation)ERM MNIST ConvNet · MNIST RS-found transformationsaccuracy: .160 -> .941Improved
add searched augmentation (ESDA) instead of random augmentation (ablation)RDA · MNIST to SVHNaccuracy: .395 -> .489Improved
add searched augmentation (ESDA) instead of random augmentation (ablation)RDA · MNIST to MNIST-Maccuracy: .751 -> .816Improved
add searched augmentation (ESDA) to ERM (ablation)ERM · MNIST to USPSaccuracy: .812 -> .840Improved
add searched augmentation (RSDA) instead of ERM (ablation)ERM WRN-16-4 · CIFAR-10 to STLaccuracy: .466 -> .526Improved
add searched augmentation (ESDA) instead of random augmentation (ablation)RDA WRN-16-4 · CIFAR-10 ES-found transformationsaccuracy: .862 -> .913Improved
add searched augmentation (RSDA) to ERM (ablation)ERM WRN-16-4 · CIFAR-10.1accuracy: .872 -> .886Improved
add searched augmentation (RSDA) instead of random augmentation (ablation)RDA FC-DenseNet103 · CamVid RS-found transformationspixel accuracy: .812 -> .825Improved
add robust augmentation (RDA/RSDA/ESDA) to ERM on clean dataERM MNIST ConvNet · MNIST clean testaccuracy: .993 -> .992No clear effect
add searched augmentation (RSDA) to ERM on clean dataERM FC-DenseNet103 · CamVid clean testpixel accuracy: .862 -> .854Hurt

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