ImageNet-C

ImagesCC BY 4.0Introduced 2019-01-01

ImageNet-C is an open source data set that consists of algorithmically generated corruptions (blur, noise) applied to the ImageNet test-set.

Source: Selective Brain Damage: Measuring the Disparate Impact of Model Pruning Image Source: https://arxiv.org/pdf/1807.01697.pdf

Benchmarks

Adversarial Robustness/mean Corruption Error (mCE)Domain Adaptation/mean Corruption Error (mCE)Domain Adaptation/Top 1 Accuracy Domain Adaptation/Number of params Domain Adaptation/Mean Accuracy Domain Generalization/mean Corruption Error (mCE)Domain Generalization/Top 1 Accuracy Domain Generalization/Number of params Unsupervised Domain Adaptation/mean Corruption Error (mCE)

Related Benchmarks

ImageNet-Caltech/Domain Adaptation/Accuracy (%)