AI Paper Atlas

Research map / Benchmarks / CIFAR-10 clean

CIFAR-10 clean leaderboard

31 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on CIFAR-10 clean. Best accuracy: 96.28 (Parseval, ICML 2017).

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.

CIFAR-10 clean · accuracy

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

#accuracy ↑MethodPaperTopic
196.28Parseval
Wide ResNet-28-10
Parseval Networks: Improving Robustness to Adversarial Examples
ICML 2017 · Table 1
Certified and empirical robustness
296SuperLoss
WideResNet-28-10
SuperLoss: A Generic Loss for Robust Curriculum Learning
NeurIPS 2020 · text
Noisy labels
396Logistic-classifier correction
WResNet
The Odds are Odd: A Statistical Test for Detecting Adversarial Examples
ICML 2019 · Table: Accuracies of our corre
Certified and empirical robustness
495.9SNGP
Wide ResNet-28-10
Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awareness
NeurIPS 2020 · CIFAR-10 results table
Uncertainty and Gaussian processes
595.9Normal + DQ
Wide ResNet-28-10
Defensive Quantization: When Efficiency Meets Robustness
ICLR 2019 · Table 3
Certified and empirical robustness
695.4Noisy Feature Mixup (α=1.0)
Pre-activated ResNet-18
Noisy Feature Mixup
ICLR 2022 · Table: Robustness of ResNet-18
Noisy labels
794.9ME-Net
ResNet-18
ME-Net: Towards Effective Adversarial Robustness with Matrix Estimation
ICML 2019 · Table 9
Black-box and transfer attacks
894.84AAA-linear
WideResNet-28
Adversarial Attack on Attackers: Post-Process to Mitigate Black-Box Score-Based Query Attacks
NeurIPS 2022 · Table 2
Black-box and transfer attacks
993.9JARN
Wide-ResNet
Jacobian Adversarially Regularized Networks for Robustness
ICLR 2020 · CIFAR-10 results table
Certified and empirical robustness
1091.99AGKD-BML
WRN-28-10
AGKD-BML: Defense Against Adversarial Attack by Attention Guided Knowledge Distillation and Bi-Directional Metric Learning
ICCV 2021 · Table 1
Certified and empirical robustness
21 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

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