State-of-the-art table for Image Classification on SVHN

Rank	Method	Percentage error	Paper	Year-Month
1	WRN + fixup init + mixup + cutout	1.4	Fixup Initialization: Residual Learning Without Normalization	2019-1
2	EraseReLU	1.54	EraseReLU: A Simple Way to Ease the Training of Deep Convolution Neural Networks	2017-9
3	CoPaNet-R-164	1.58	Deep Competitive Pathway Networks	2017-9
4	DenseNet	1.59	Densely Connected Convolutional Networks	2016-8
5	Tree+Max-Avg pooling	1.7	Generalizing Pooling Functions in Convolutional Neural Networks: Mixed, Gated, and Tree	2015-9
6	Wide ResNet	1.7	Wide Residual Networks	2016-5
7	Stochastic Depth	1.75	Deep Networks with Stochastic Depth	2016-3
8	CMsC	1.8	Competitive Multi-scale Convolution	2015-11
9	RCNN-96	1.8		-
10	BNM NiN	1.8	Batch-normalized Maxout Network in Network	2015-11
11	DSN	1.9	Deeply-Supervised Nets	2014-9
12	MLR DNN	1.9		-
13	Regularization of Neural Networks using DropConnect	1.9		-
14	MIM	2.0	On the Importance of Normalisation Layers in Deep Learning with Piecewise Linear Activation Units	2015-8
15	FractalNet	2.01	FractalNet: Ultra-Deep Neural Networks without Residuals	2016-5
16	BinaryConnect	2.2	BinaryConnect: Training Deep Neural Networks with binary weights during propagations	-
17	DCNN	2.2	Multi-digit Number Recognition from Street View Imagery using Deep Convolutional Neural Networks	2013-12
18	Network in Network	2.35	Network In Network	2013-12
19	ReNet	2.4	ReNet: A Recurrent Neural Network Based Alternative to Convolutional Networks	2015-5
20	Maxout	2.5	Maxout Networks	2013-2
21	Stochastic Pooling	2.8	Stochastic Pooling for Regularization of Deep Convolutional Neural Networks	2013-1
22	Deep Complex	3.3	Deep Complex Networks	2017-5
23	FLSCNN	4.0	Enhanced Image Classification With a Fast-Learning Shallow Convolutional Neural Network	2015-3
24	Convolutional neural networks applied to house numbers digit classification	4.9		-
25	CLS-GAN	5.98	Loss-Sensitive Generative Adversarial Networks on Lipschitz Densities	2017-1

Rank	Method	Percentage error	Paper title	Year
1	WRN + fixup init + mixup + cutout	1.4	Fixup Initialization: Residual Learning Without Normalization	2019
2	EraseReLU	1.54	EraseReLU: A Simple Way to Ease the Training of Deep Convolution Neural Networks	2017
3	CoPaNet-R-164	1.58	Deep Competitive Pathway Networks	2017
4	DenseNet	1.59	Densely Connected Convolutional Networks	2016
5	Tree+Max-Avg pooling	1.7	Generalizing Pooling Functions in Convolutional Neural Networks: Mixed, Gated, and Tree	2015
6	Wide ResNet	1.7	Wide Residual Networks	2016
7	Stochastic Depth	1.75	Deep Networks with Stochastic Depth	2016
8	CMsC	1.8	Competitive Multi-scale Convolution	2015
9	RCNN-96	1.8
10	BNM NiN	1.8	Batch-normalized Maxout Network in Network	2015
11	DSN	1.9	Deeply-Supervised Nets	2014
12	MLR DNN	1.9
13	Regularization of Neural Networks using DropConnect	1.9
14	MIM	2.0	On the Importance of Normalisation Layers in Deep Learning with Piecewise Linear Activation Units	2015
15	FractalNet	2.01	FractalNet: Ultra-Deep Neural Networks without Residuals	2016
16	BinaryConnect	2.2	BinaryConnect: Training Deep Neural Networks with binary weights during propagations
17	DCNN	2.2	Multi-digit Number Recognition from Street View Imagery using Deep Convolutional Neural Networks	2013
18	Network in Network	2.35	Network In Network	2013
19	ReNet	2.4	ReNet: A Recurrent Neural Network Based Alternative to Convolutional Networks	2015
20	Maxout	2.5	Maxout Networks	2013
21	Stochastic Pooling	2.8	Stochastic Pooling for Regularization of Deep Convolutional Neural Networks	2013
22	Deep Complex	3.3	Deep Complex Networks	2017
23	FLSCNN	4.0	Enhanced Image Classification With a Fast-Learning Shallow Convolutional Neural Network	2015
24	Convolutional neural networks applied to house numbers digit classification	4.9
25	CLS-GAN	5.98	Loss-Sensitive Generative Adversarial Networks on Lipschitz Densities	2017

Image Classification on SVHN Edit Add Remove

Image Classification on SVHN

Add Remove