Image Classification on MNIST

Metric: Percentage error (lower is better)

LeaderboardDataset

Loading chart...

Results

Sort:

#	Model↕	Percentage error▲	Extra Data	Paper	Date↕	Code
1	Branching/Merging CNN + Homogeneous Vector Capsules	0.13	No	No Routing Needed Between Capsules	2020-01-24	Code
2	EnsNet (Ensemble learning in CNN augmented with fully connected subnetworks)	0.16	No	Ensemble learning in CNN augmented with fully co...	2020-03-19	Code
3	Efficient-CapsNet	0.16	No	Efficient-CapsNet: Capsule Network with Self-Att...	2021-01-29	Code
4	SOPCNN (Only a single Model)	0.17	No	Stochastic Optimization of Plain Convolutional N...	2020-01-24	Code
5	RMDL (30 RDLs)	0.18	No	RMDL: Random Multimodel Deep Learning for Classi...	2018-05-03	Code
6	R-ExplaiNet-22 (single model)	0.2	No	Learning local discrete features in explainable-...	2024-10-31	Code
7	DropConnect	0.21	No	-	-	Code
8	MCDNN	0.23	No	Multi-column Deep Neural Networks for Image Clas...	2012-02-13	Code
9	APAC	0.23	No	APAC: Augmented PAttern Classification with Neur...	2015-05-13	-
10	BNM NiN	0.24	No	Batch-normalized Maxout Network in Network	2015-11-09	Code
11	SimpleNetv1	0.25	No	Lets keep it simple, Using simple architectures ...	2016-08-22	Code
12	CapsNet	0.25	No	Dynamic Routing Between Capsules	2017-10-26	Code
13	VGG8B + LocalLearning + CO	0.26	No	Training Neural Networks with Local Error Signals	2019-01-20	Code
14	VGG-5 (Spinal FC)	0.28	No	SpinalNet: Deep Neural Network with Gradual Input	2020-07-07	Code
15	TextCaps	0.29	No	TextCaps : Handwritten Character Recognition wit...	2019-04-17	Code
16	ExquisiteNetV2	0.29	No	A Novel lightweight Convolutional Neural Network...	2021-05-19	Code
17	WaveMix-128/7	0.29	No	WaveMix: Resource-efficient Token Mixing for Ima...	2022-03-07	Code
18	Fractional MP	0.3	No	Fractional Max-Pooling	2014-12-18	Code
19	Tree+Max-Avg pooling	0.3	No	Generalizing Pooling Functions in Convolutional ...	2015-09-30	Code
20	CMsC	0.3	No	Competitive Multi-scale Convolution	2015-11-18	-
21	EXACT (M3-CNN)	0.33	No	EXACT: How to Train Your Accuracy	2022-05-19	Code
22	Second Order Neural Ordinary Differential Equation	0.37	No	On Second Order Behaviour in Augmented Neural ODEs	2020-06-12	Code
23	Augmented Neural Ordinary Differential Equation	0.37	No	Augmented Neural ODEs	2019-04-02	Code
24	DSN	0.4	No	Deeply-Supervised Nets	2014-09-18	Code
25	CKN	0.4	No	Convolutional Kernel Networks	2014-06-12	-
26	C-SVDDNet	0.4	No	Unsupervised Feature Learning with C-SVDDNet	2014-12-23	-
27	HOPE	0.4	No	Hybrid Orthogonal Projection and Estimation (HOP...	2015-02-03	-
28	FLSCNN	0.4	No	Enhanced Image Classification With a Fast-Learni...	2015-03-16	-
29	MIM	0.4	No	On the Importance of Normalisation Layers in Dee...	2015-08-03	-
30	Fitnet-LSUV-SVM	0.4	No	All you need is a good init	2015-11-19	Code
31	Neural Architecture Search (NAS)-enabled Convolutional Neural Network (CNN)	0.5	No	-	-	-
32	Maxout Networks	0.5	No	Maxout Networks	2013-02-18	Code
33	NiN	0.5	No	Network In Network	2013-12-16	Code
34	ReNet	0.5	No	ReNet: A Recurrent Neural Network Based Alternat...	2015-05-03	Code
35	DCNN+GFE	0.5	No	Deep Convolutional Neural Networks as Generic Fe...	2017-10-06	-
36	VDN	0.5	No	Training Very Deep Networks	2015-07-22	Code
37	NeuPDE	0.51	No	NeuPDE: Neural Network Based Ordinary and Partia...	2019-08-08	-
38	Simple CNN with BaikalCMA loss	0.53	No	Improved Training Speed, Accuracy, and Data Util...	2019-05-27	Code
39	SEER (RegNet10B)	0.58	No	Vision Models Are More Robust And Fair When Pret...	2022-02-16	Code
40	Convolutional Tsetlin Machine	0.6	No	The Convolutional Tsetlin Machine	2019-05-23	Code
41	PCANet	0.6	No	PCANet: A Simple Deep Learning Baseline for Imag...	2014-04-14	Code
42	DiffPrune (LeNet5)	0.6	No	DiffPrune: Neural Network Pruning with Determini...	2020-12-07	Code
43	Deep Fried Convnets	0.7	No	Deep Fried Convnets	2014-12-22	Code
44	Sparse Activity and Sparse Connectivity in Supervised Learning	0.8	No	Sparse Activity and Sparse Connectivity in Super...	2016-03-28	-
45	Explaining and Harnessing Adversarial Examples	0.8	No	Explaining and Harnessing Adversarial Examples	2014-12-20	Code
46	BinaryConnect	1	No	BinaryConnect: Training Deep Neural Networks wit...	2015-11-02	Code
47	Convolutional PMM (Parametric Matrix Model)	1.01	No	Parametric Matrix Models	2024-01-22	-
48	LeNet 300-100 (Sparse Momentum)	1.26	No	Sparse Networks from Scratch: Faster Training wi...	2019-07-10	Code
49	Convolutional Clustering	1.4	No	Convolutional Clustering for Unsupervised Learning	2015-11-19	-
50	CNN Model by Som	1.41	No	Convolutional Sequence to Sequence Learning	2017-05-08	Code
51	Weighted Tsetlin Machine	1.5	No	The Weighted Tsetlin Machine: Compressed Represe...	2019-11-28	Code
52	MLP (ideal number of groups)	1.67	No	On the Ideal Number of Groups for Isometric Grad...	2023-02-07	-
53	Perceptron with a tensor train layer	1.8	No	Tensorizing Neural Networks	2015-09-22	Code
54	ANODE	1.8	No	Augmented Neural ODEs	2019-04-02	Code
55	Tsetlin Machine	1.8	No	The Tsetlin Machine - A Game Theoretic Bandit Dr...	2018-04-04	Code
56	GECCO	1.96	No	A Single Graph Convolution Is All You Need: Effi...	2024-02-01	Code
57	PMM (Parametric Matrix Model)	2.62	No	Parametric Matrix Models	2024-01-22	-
58	DNN-5 (Trainable Activations)	2.8	No	-	-	Code
59	DNN-3 (Trainable Activations)	3	No	-	-	Code
60	DNN-2 (Trainable Activations)	3.6	No	-	-	Code
61	Zhao et al. (2015) (auto-encoder)	4.76	No	Stacked What-Where Auto-encoders	2015-06-08	Code
62	ProjectionNet	5	No	ProjectionNet: Learning Efficient On-Device Deep...	2017-08-02	-

#1Branching/Merging CNN + Homogeneous Vector CapsulesSOTA
0.13
Percentage error· 2020-01-24
No Routing Needed Between Capsules Code
#2EnsNet (Ensemble learning in CNN augmented with fully connected subnetworks)
0.16
Percentage error· 2020-03-19
Ensemble learning in CNN augmented with fully connected subnetworks Code
#3Efficient-CapsNet
0.16
Percentage error· 2021-01-29
Efficient-CapsNet: Capsule Network with Self-Attention Routing Code
#4SOPCNN (Only a single Model)SOTA
0.17
Percentage error· 2020-01-24
Stochastic Optimization of Plain Convolutional Neural Networks with Simple methods Code
#5RMDL (30 RDLs)SOTA
0.18
Percentage error· 2018-05-03
RMDL: Random Multimodel Deep Learning for Classification Code
#6R-ExplaiNet-22 (single model)
0.2
Percentage error· 2024-10-31
Learning local discrete features in explainable-by-design convolutional neural networks Code
#7DropConnect
0.21
Percentage error
No paperCode
#8MCDNNSOTA
0.23
Percentage error· 2012-02-13
Multi-column Deep Neural Networks for Image Classification Code
#9APAC
0.23
Percentage error· 2015-05-13
APAC: Augmented PAttern Classification with Neural Networks
#10BNM NiN
0.24
Percentage error· 2015-11-09
Batch-normalized Maxout Network in Network Code
#11SimpleNetv1
0.25
Percentage error· 2016-08-22
Lets keep it simple, Using simple architectures to outperform deeper and more complex architectures Code
#12CapsNet
0.25
Percentage error· 2017-10-26
Dynamic Routing Between Capsules Code
#13VGG8B + LocalLearning + CO
0.26
Percentage error· 2019-01-20
Training Neural Networks with Local Error Signals Code
#14VGG-5 (Spinal FC)
0.28
Percentage error· 2020-07-07
SpinalNet: Deep Neural Network with Gradual Input Code
#15TextCaps
0.29
Percentage error· 2019-04-17
TextCaps : Handwritten Character Recognition with Very Small Datasets Code
#16ExquisiteNetV2
0.29
Percentage error· 2021-05-19
A Novel lightweight Convolutional Neural Network, ExquisiteNetV2 Code
#17WaveMix-128/7
0.29
Percentage error· 2022-03-07
WaveMix: Resource-efficient Token Mixing for Images Code
#18Fractional MP
0.3
Percentage error· 2014-12-18
Fractional Max-Pooling Code
#19Tree+Max-Avg pooling
0.3
Percentage error· 2015-09-30
Generalizing Pooling Functions in Convolutional Neural Networks: Mixed, Gated, and Tree Code
#20CMsC
0.3
Percentage error· 2015-11-18
Competitive Multi-scale Convolution
#21EXACT (M3-CNN)
0.33
Percentage error· 2022-05-19
EXACT: How to Train Your Accuracy Code
#22Second Order Neural Ordinary Differential Equation
0.37
Percentage error· 2020-06-12
On Second Order Behaviour in Augmented Neural ODEs Code
#23Augmented Neural Ordinary Differential Equation
0.37
Percentage error· 2019-04-02
Augmented Neural ODEs Code
#24DSN
0.4
Percentage error· 2014-09-18
Deeply-Supervised Nets Code
#25CKN
0.4
Percentage error· 2014-06-12
Convolutional Kernel Networks
#26C-SVDDNet
0.4
Percentage error· 2014-12-23
Unsupervised Feature Learning with C-SVDDNet
#27HOPE
0.4
Percentage error· 2015-02-03
Hybrid Orthogonal Projection and Estimation (HOPE): A New Framework to Probe and Learn Neural Networks
#28FLSCNN
0.4
Percentage error· 2015-03-16
Enhanced Image Classification With a Fast-Learning Shallow Convolutional Neural Network
#29MIM
0.4
Percentage error· 2015-08-03
On the Importance of Normalisation Layers in Deep Learning with Piecewise Linear Activation Units
#30Fitnet-LSUV-SVM
0.4
Percentage error· 2015-11-19
All you need is a good init Code
#31Neural Architecture Search (NAS)-enabled Convolutional Neural Network (CNN)
0.5
Percentage error
No paper
#32Maxout Networks
0.5
Percentage error· 2013-02-18
Maxout Networks Code
#33NiN
0.5
Percentage error· 2013-12-16
Network In Network Code
#34ReNet
0.5
Percentage error· 2015-05-03
ReNet: A Recurrent Neural Network Based Alternative to Convolutional Networks Code
#35DCNN+GFE
0.5
Percentage error· 2017-10-06
Deep Convolutional Neural Networks as Generic Feature Extractors
#36VDN
0.5
Percentage error· 2015-07-22
Training Very Deep Networks Code
#37NeuPDE
0.51
Percentage error· 2019-08-08
NeuPDE: Neural Network Based Ordinary and Partial Differential Equations for Modeling Time-Dependent Data
#38Simple CNN with BaikalCMA loss
0.53
Percentage error· 2019-05-27
Improved Training Speed, Accuracy, and Data Utilization Through Loss Function Optimization Code
#39SEER (RegNet10B)
0.58
Percentage error· 2022-02-16
Vision Models Are More Robust And Fair When Pretrained On Uncurated Images Without Supervision Code
#40Convolutional Tsetlin Machine
0.6
Percentage error· 2019-05-23
The Convolutional Tsetlin Machine Code
#41PCANet
0.6
Percentage error· 2014-04-14
PCANet: A Simple Deep Learning Baseline for Image Classification?Code
#42DiffPrune (LeNet5)
0.6
Percentage error· 2020-12-07
DiffPrune: Neural Network Pruning with Deterministic Approximate Binary Gates and $L_0$ Regularization Code
#43Deep Fried Convnets
0.7
Percentage error· 2014-12-22
Deep Fried Convnets Code
#44Sparse Activity and Sparse Connectivity in Supervised Learning
0.8
Percentage error· 2016-03-28
Sparse Activity and Sparse Connectivity in Supervised Learning
#45Explaining and Harnessing Adversarial Examples
0.8
Percentage error· 2014-12-20
Explaining and Harnessing Adversarial Examples Code
#46BinaryConnect
1
Percentage error· 2015-11-02
BinaryConnect: Training Deep Neural Networks with binary weights during propagations Code
#47Convolutional PMM (Parametric Matrix Model)
1.01
Percentage error· 2024-01-22
Parametric Matrix Models
#48LeNet 300-100 (Sparse Momentum)
1.26
Percentage error· 2019-07-10
Sparse Networks from Scratch: Faster Training without Losing Performance Code
#49Convolutional Clustering
1.4
Percentage error· 2015-11-19
Convolutional Clustering for Unsupervised Learning
#50CNN Model by Som
1.41
Percentage error· 2017-05-08
Convolutional Sequence to Sequence Learning Code
#51Weighted Tsetlin Machine
1.5
Percentage error· 2019-11-28
The Weighted Tsetlin Machine: Compressed Representations with Weighted Clauses Code
#52MLP (ideal number of groups)
1.67
Percentage error· 2023-02-07
On the Ideal Number of Groups for Isometric Gradient Propagation
#53Perceptron with a tensor train layer
1.8
Percentage error· 2015-09-22
Tensorizing Neural Networks Code
#54ANODE
1.8
Percentage error· 2019-04-02
Augmented Neural ODEs Code
#55Tsetlin Machine
1.8
Percentage error· 2018-04-04
The Tsetlin Machine - A Game Theoretic Bandit Driven Approach to Optimal Pattern Recognition with Propositional Logic Code
#56GECCO
1.96
Percentage error· 2024-02-01
A Single Graph Convolution Is All You Need: Efficient Grayscale Image Classification Code
#57PMM (Parametric Matrix Model)
2.62
Percentage error· 2024-01-22
Parametric Matrix Models
#58DNN-5 (Trainable Activations)
2.8
Percentage error
No paperCode
#59DNN-3 (Trainable Activations)
3
Percentage error
No paperCode
#60DNN-2 (Trainable Activations)
3.6
Percentage error
No paperCode
#61Zhao et al. (2015) (auto-encoder)
4.76
Percentage error· 2015-06-08
Stacked What-Where Auto-encoders Code
#62ProjectionNet
5
Percentage error· 2017-08-02
ProjectionNet: Learning Efficient On-Device Deep Networks Using Neural Projections