ResearchBenchmarksRobustness Evaluation on ImageNet Robustness Variants (IN-A, IN-R, IN-Sketch, IN-C) (test)For AI assistantsMarkdownContent indexRobustness Evaluation on ImageNet Robustness Variants (IN-A, IN-R, IN-Sketch, IN-C) (test)Follow43.8IN-A Top-1 AccSiameseIM31.94435.02238.141.178Jun 2, 2022Evaluation resultsMethodMethodLinksIN-A Top-1 AccIN-R Top-1 AccIN-Sketch Top-1 AccIN-C mCESiameseIMEpochs=1600Epochs=16002022.0643.852.538.357.1SiameseIMEpochs=400Epochs=4002022.0638.651.637.755.9MSNEpochs=1200, multi-cro...Epochs=1200, multi-crop=true, symmetric loss=true2022.0637.55036.353.4MAEEpochs=1600Epochs=16002022.0635.948.334.548.3MoCo-v3Epochs=600, symmetric...Epochs=600, symmetric loss=true2022.0632.449.835.955.4