AI Paper Atlas

Research map / Benchmarks / ImageNet-O

ImageNet-O leaderboard

10 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on ImageNet-O. Best AUPR: 54.61 (ViT B/16, ImageNet-21k pre-training, AAAI 2022).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

ImageNet-O · AUPR

10 papers, each with its best reported number for its own method; higher is better.

#AUPR ↑MethodPaperTopic
154.61ViT B/16, ImageNet-21k pre-training
ViT B/16
Vision Transformers Are Robust Learners
AAAI 2022 · Pre-training results table
Attention and transformers
231.14Dependency regularization
ViT-S
Neural Dependencies Emerging From Learning Massive Categories
CVPR 2023 · Table: Metrics of using and no
Recognition and long-tail
328.42ViT-SPKDEDesigning Robust Transformers using Robust Kernel Density Estimation
NeurIPS 2023 · Table 3
Certified and empirical robustness
422.6DeiT-S + MJP
DeiT-S
Masked Jigsaw Puzzle: A Versatile Position Embedding for Vision Transformers
CVPR 2023 · Table 2
Attention and transformers
521.9TransMix-DeiT-S
DeiT-S
TransMix: Attend To Mix for Vision Transformers
CVPR 2022 · Table 4
CNN features
620.35RPC-SymViT (5iter/layer1)
ViT-tiny
Unveiling the Hidden Structure of Self-Attention via Kernel Principal Component Analysis
NeurIPS 2024 · Table 1
Attention and transformers
719.55Swin-ACMoE-Top 2
Swin Transformer
Tight Clusters Make Specialized Experts
ICLR 2025 · Table: Test Accuracy on Image
Mixture of experts
819.22PID DeiT
DeiT-tiny
PIDformer: Transformer Meets Control Theory
ICML 2024 · Table: Evaluation of PID DeiT
Attention and transformers
918.2IPMix
ResNet-50
IPMix: Label-Preserving Data Augmentation Method for Training Robust Classifiers
NeurIPS 2023 · Table 3
Certified and empirical robustness
1016.69Catch-up Mix
ResNet-50
Catch-Up Mix: Catch-Up Class for Struggling Filters in CNN
AAAI 2024 · Table 5
Recognition and long-tail

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.