Out Of Distribution Detection On Imagenet 1K 10

Metrics

AUROC

FPR95

Results

Performance results of various models on this benchmark

			Paper Title
RP+GradNorm	-	79.80	Detecting Out-of-distribution Data through In-distribution Class Prior
GEN	-	77.93	GEN: Pushing the Limits of Softmax-Based Out-of-Distribution Detection
Watermarking (WRN-40-2 w/ Energy)	80.80	67.75	Watermarking for Out-of-distribution Detection
GradNorm (ResNetv2-101)	-	61.42	On the Importance of Gradients for Detecting Distributional Shifts in the Wild
Watermarking (WRN-40-2 w/ MSP)	84.00	61.2	Watermarking for Out-of-distribution Detection
MOS (BiT-S-R101x1)	81.23	60.43	MOS: Towards Scaling Out-of-distribution Detection for Large Semantic Space
MCM (CLIP-L)	84.88	59.88	Delving into Out-of-Distribution Detection with Vision-Language Representations
MCM (CLIP-B)	86.11	57.77	Delving into Out-of-Distribution Detection with Vision-Language Representations
SSD	85.4	57.2	SSD: A Unified Framework for Self-Supervised Outlier Detection
DML	-	52.80	Decoupling MaxLogit for Out-of-Distribution Detection
NPOS	88.80	46.12	Non-Parametric Outlier Synthesis
SHE	-	45.09	Out-of-Distribution Detection based on In-Distribution Data Patterns Memorization with Modern Hopfield Energy
ODIN+UMAP (ResNet-50)	88.35	42.02	Unleashing Mask: Explore the Intrinsic Out-of-Distribution Detection Capability
BATS (ResNet-50)	92.27	38.9	Boosting Out-of-distribution Detection with Typical Features
ReAct (ResNet-50)	91.53	38.42	ReAct: Out-of-distribution Detection With Rectified Activations
RankFeat (ResNetv2-101)	91.7	37.29	RankFeat: Rank-1 Feature Removal for Out-of-distribution Detection
DOE	88.9	34.67	Out-of-distribution Detection with Implicit Outlier Transformation
DICE (ResNet-50)	90.3	31.72	DICE: Leveraging Sparsification for Out-of-Distribution Detection
DICE + ReAct (ResNet-50)	92.74	28.07	DICE: Leveraging Sparsification for Out-of-Distribution Detection
LINe (ResNet-50)	94.44	22.54	LINe: Out-of-Distribution Detection by Leveraging Important Neurons

0 of 33 row(s) selected.

Command Palette

Out Of Distribution Detection On Imagenet 1K 10

Metrics

Results