ResearchBenchmarksOpen-Set Recognition on ImageNet Semantic Shift Benchmark HardFor AI assistantsMarkdownContent indexOpen-Set Recognition on ImageNet Semantic Shift Benchmark HardFollow78.8AccuracyMLS78.17678.33878.578.662Oct 12, 2021Evaluation resultsMethodMethodLinksAccuracyAUROCOSCRMLSBackbone=ResNet-50, Pr...Backbone=ResNet-50, Pre-training=ImageNet-1K2021.1078.872.863.4ARPL+Backbone=ResNet-50, Pr...Backbone=ResNet-50, Pre-training=ImageNet-1K2021.1078.27463