AI Paper Atlas

Research map / Benchmarks / SVHN

SVHN leaderboard

224 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on SVHN. Best accuracy: 99.64 (Lineage Detector, CVPR 2024).

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.

SVHN · accuracy

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

#accuracy ↑MethodPaperTopic
199.64Lineage Detector
ResNet18
Neural Lineage
CVPR 2024 · Table 1
Open-set and few-shot detection
299RandAugment
Wide-ResNet-28-10
RandAugment: Practical Automated Data Augmentation with a Reduced Search Space
NeurIPS 2020 · Table 1, CIFAR results
CNN features
399ResNeXt-50 + CBS
ResNeXt-50
Curriculum By Smoothing
NeurIPS 2020 · Table 1
CNN design
498.6Noisy Student Training (B0)
EfficientNet-B0
Self-Training With Noisy Student Improves ImageNet Classification
CVPR 2020 · Table: Results on SVHN
Knowledge distillation
598.36JumpoutJumpout : Improved Dropout for Deep Neural Networks with ReLUs
ICML 2019 · Table 2
RNNs and normalization
698.3DivAug
Wide-ResNet-28-10
DivAug: Plug-In Automated Data Augmentation With Explicit Diversity Maximization
ICCV 2021 · Table 2
Semi-supervised learning
798.3Ours
Wide-ResNet-28-10
On the Generalization Effects of Linear Transformations in Data Augmentation
ICML 2020 · Table 3
Recognition and long-tail
898.3LPOMLifted Proximal Operator Machines
AAAI 2019 · Table 2
Implicit bias and generalization
998.27KSE (G=4)
DenseNet-40-12
Exploiting Kernel Sparsity and Entropy for Interpretable CNN Compression
CVPR 2019 · Table 3
CNN design
1098.1BORAT3
WRN 16-4
A Stochastic Bundle Method for Interpolation
JMLR 2022 · Table 1
Nonconvex and smooth optimization
117 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

SVHN · error rate

57 papers, each with its best reported number for its own method; lower is better.

#error rate ↓MethodPaperTopic
11.18PBA
Wide-ResNet-28-10
Population Based Augmentation: Efficient Learning of Augmentation Policy Schedules
ICML 2019 · Table 2
Visual and representation RL
21.2DADA
Wide-ResNet-28-10
DADA: Differentiable Automatic Data Augmentation
ECCV 2020 · Table 3
Neural architecture search
31.2Faster AA
WideResNet-28-10
Faster AutoAugment: Learning Augmentation Strategies Using Backpropagation
ECCV 2020 · Table 2
Neural architecture search
41.4Fixup-init + Mixup + Cutout
WideResNet-16-12
Fixup Initialization: Residual Learning Without Normalization
ICLR 2019 · Additional CIFAR-10, SVHN tabl
RNNs and normalization
51.48WRN-28-SORTSORT: Second-Order Response Transform for Visual Recognition
ICCV 2017 · Table 2
CNN design
61.51Snapshot Ensemble (α₀=0.2)
Wide-ResNet-32
Snapshot Ensembles: Train 1, Get M for Free
ICLR 2017 · Table 1
RNNs and normalization
71.52EncapNet++Neural Network Encapsulation
ECCV 2018 · Table 4
Pruning and efficient tuning
81.53CliqueNet (k=80, T=15)Convolutional Neural Networks With Alternately Updated Clique
CVPR 2018 · Table (CIFAR and SVHN results)
CNN design
91.54CapProNet+WideResNet with Dropout (c=8)
WideResNet-16
CapProNet: Deep Feature Learning via Orthogonal Projections onto Capsule Subspaces
NeurIPS 2018 · Table 1
Recognition and long-tail
101.54OrthoReg on Wide ResNet
Wide ResNet
Regularizing CNNs with Locally Constrained Decorrelations
ICLR 2017 · Table 4
CNN design
47 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

SVHN · FID

20 papers, each with its best reported number for its own method; lower is better.

#FID ↓MethodPaperTopic
13.87LA–AltGANTaming GANs with Lookahead-Minmax
ICLR 2021 · Table 1
Optimizers
24.38Our KD
Hierarchical VAE, 0.10M parameters
A Unified Knowledge Distillation Framework for Deep Directed Graphical Models
CVPR 2023 · Table 1
Variational and generative sampling
34.81SVRE
Deep ResNet
Reducing Noise in GAN Training with Variance Reduced Extragradient
NeurIPS 2019 · Table (best obtained FID score
Nonconvex and smooth optimization
410.07Progressive Coding HVAE
VDVAE
Bit Prioritization in Variational Autoencoders via Progressive Coding
ICML 2022 · Table 2
Autoregressive and latent generation
512.5TTUR
DCGAN
GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
NeurIPS 2017 · Table 1
Optimizers
615.32CoopFlow(Pre)A Tale of Two Flows: Cooperative Learning of Langevin Flow and Normalizing Flow Toward Energy-Based Model
ICLR 2022 · Table: FID scores on CIFAR-10,
Variational and generative sampling
718.76Ours-DAMCLearning Energy-Based Prior Model with Diffusion-Amortized MCMC
NeurIPS 2023 · Table 1
Fast diffusion sampling
819.48OursLearning Deep Latent Variable Models by Short-Run MCMC Inference With Optimal Transport Correction
CVPR 2021 · Table 1
Variational and generative sampling
920.19FCE
Glow
Flow Contrastive Estimation of Energy-Based Models
CVPR 2020 · Table 1
Semi-supervised learning
1023.64LFBM-MCMCA Tale of Two Latent Flows: Learning Latent Space Normalizing Flow with Short-Run Langevin Flow for Approximate Inference
AAAI 2023 · Image reconstruction and gener
Variational and generative sampling
10 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

SVHN · AUROC

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

1 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

SVHN · clean accuracy (%)

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

#clean accuracy (%) ↑MethodPaperTopic
195.59AVmixup
WRN-16-8
Adversarial Vertex Mixup: Toward Better Adversarially Robust Generalization
CVPR 2020 · Table: comparisons on other da
Certified and empirical robustness
295PRL
ResNet-18
Probabilistically Robust Learning: Balancing Average and Worst-case Performance
ICML 2022 · Table 2
Certified and empirical robustness
393.98adv-comp-sum
WRN-34-20
Cross-Entropy Loss Functions: Theoretical Analysis and Applications
ICML 2023 · Table 2
Prediction and decision losses
493.7Multi-exit (NEED)
VGG-16 (5 exits)
Rethinking the Adversarial Robustness of Multi-Exit Neural Networks in an Attack-Defense Game
CVPR 2025 · Table 2
Certified and empirical robustness
593Ours
ResNet18
Generalization Bound and New Algorithm for Clean-Label Backdoor Attack
ICML 2024 · Evaluations on more datasets
Backdoor and poisoning
692.8MAD
VGG-16
Masking Adversarial Damage: Finding Adversarial Saliency for Robust and Sparse Network
CVPR 2022 · Table 1
Certified and empirical robustness
791.1CURE
ResNet-18
Robustness via Curvature Regularization, and Vice Versa
CVPR 2019 · Appendix: SVHN results table
Certified and empirical robustness
890.68ST
ResNet-18
Squeeze Training for Adversarial Robustness
ICLR 2023 · Table 1
Certified and empirical robustness
989.62TPAP
ResNet-18
Robust Overfitting Does Matter: Test-Time Adversarial Purification With FGSM
CVPR 2024 · Table 1
Certified and empirical robustness
1082.9ATDAImproving the Generalization of Adversarial Training with Domain Adaptation
ICLR 2019 · Table 1
Certified and empirical robustness

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