HyperAIHyperAI超神经

Command Palette

Search for a command to run...

算力平台

Out Of Distribution Detection On Imagenet 1K 9

评估指标

AUROC
FPR95

评测结果

各个模型在此基准测试上的表现结果

Paper Title
RP+GradNorm-83.29Detecting Out-of-distribution Data through In-distribution Class Prior
KNN (ResNet-50)74.8777.09Out-of-Distribution Detection with Deep Nearest Neighbors
Watermarking (WRN-40-2 w/ Energy)79.8571.85Watermarking for Out-of-distribution Detection
Watermarking (WRN-40-2 w/ MSP)82.0370.59Watermarking for Out-of-distribution Detection
DOE83.0567.84Out-of-distribution Detection with Implicit Outlier Transformation
DML-61.43Decoupling MaxLogit for Out-of-Distribution Detection
GradNorm (ResNetv2-101)-60.86On the Importance of Gradients for Detecting Distributional Shifts in the Wild
KNN (ResNet-50 SupCon)84.6260.02Out-of-Distribution Detection with Deep Nearest Neighbors
ReAct (ResNet-50)86.6451.56ReAct: Out-of-distribution Detection With Rectified Activations
ODIN+UMAP (ResNet-50)86.9950.06Unleashing Mask: Explore the Intrinsic Out-of-Distribution Detection Capability
MOS (BiT-S-R101x1)89.0649.54MOS: Towards Scaling Out-of-distribution Detection for Large Semantic Space
DICE (ResNet-50)87.4846.49DICE: Leveraging Sparsification for Out-of-Distribution Detection
SHE-45.35Out-of-Distribution Detection based on In-Distribution Data Patterns Memorization with Modern Hopfield Energy
NPOS89.4445.27Non-Parametric Outlier Synthesis
MCM (CLIP-B)89.7744.69Delving into Out-of-Distribution Detection with Vision-Language Representations
ASH-S (ResNet-50)90.9839.78Extremely Simple Activation Shaping for Out-of-Distribution Detection
RankFeat (ResNetv2-101)90.9339.34RankFeat: Rank-1 Feature Removal for Out-of-distribution Detection
DICE + ReAct (ResNet-50)90.6736.86DICE: Leveraging Sparsification for Out-of-Distribution Detection
MCM (CLIP-L)92.0035.42Delving into Out-of-Distribution Detection with Vision-Language Representations
SCALE (ResNet50)92.2634.51Scaling for Training Time and Post-hoc Out-of-distribution Detection Enhancement
0 of 24 row(s) selected.