Image Classification on MNIST

92개 결과 · ⬇ CSV · JSON

Percentage error 낮을수록 좋음

Rank	Model	Percentage error	Accuracy	Trainable Parameters	Cross Entropy Loss	Epochs	Top 1 Accuracy	Paper	Code	Year
1	Branching/Merging CNN + Homogeneous Vector Capsules	0.13	99.87	1514187	–	–	–	No Routing Needed Between Capsules	AdamByerly/BMCNNwHFCs	2020
2	EnsNet (Ensemble learning in CNN augmented with fully connected subnetworks)	0.16	99.84	–	–	–	–	Ensemble learning in CNN augmented with fully connected subnetworks	eaguaida/TF_EnsNet- · dslisleedh/EnsNet-tensorflow2	2020
2	Efficient-CapsNet	0.16	99.84	161824	–	–	–	Efficient-CapsNet: Capsule Network with Self-Attention Routing	EscVM/Efficient-CapsNet · kaparoo/Efficient-CapsNet	2021
4	SOPCNN (Only a single Model)	0.17	99.83	1400000	–	–	–	Stochastic Optimization of Plain Convolutional Neural Networks with Simple methods	junaidaliop/MNIST-SOPCNN	2020
5	RMDL (30 RDLs)	0.18	99.82	–	–	–	–	RMDL: Random Multimodel Deep Learning for Classification	kk7nc/RMDL	2018
6	R-ExplaiNet-22 (single model)	0.20	99.80	743882	–	–	–	Learning local discrete features in explainable-by-design convolutional neural networks	pikaplan/LearnExplaiNet	2024
7	DropConnect	0.21	99.77	–	–	–	–	Regularization of Neural Networks using DropConnect	mlpack/mlpack	2013
8	MCDNN	0.23	–	–	–	–	–	Multi-column Deep Neural Networks for Image Classification	hughperkins/DeepCL	2012
8	APAC	0.23	–	–	–	–	–	APAC: Augmented PAttern Classification with Neural Networks		2015
10	BNM NiN	0.24	–	–	–	–	–	Batch-normalized Maxout Network in Network	JohnBensen1000/machine_learning	2015
11	SimpleNetv1	0.25	–	–	–	–	–	Lets keep it simple, Using simple architectures to outperform deeper and more complex architectures	Coderx7/SimpleNet · Coderx7/SimpleNet_Pytorch · JavierAntoran/moby_dick_whale_audio_detection · +6	2016
11	CapsNet	0.25	–	–	–	–	–	Dynamic Routing Between Capsules	labmlai/annotated_deep_learning_paper_implementations · naturomics/CapsNet-Tensorflow · Sarasra/models · +74	2017
13	VGG8B + LocalLearning + CO	0.26	–	–	–	–	–	Training Neural Networks with Local Error Signals	anokland/local-loss · ai-tech-research-lab/nitro-d	2019
14	VGG-5 (Spinal FC)	0.28	99.72	–	–	–	–	SpinalNet: Deep Neural Network with Gradual Input	dipuk0506/SpinalNet · dipuk0506/uq · Mechachleopteryx/SpinalNet	2020
15	TextCaps	0.29	99.71	–	–	–	–	TextCaps : Handwritten Character Recognition with Very Small Datasets	vinojjayasundara/textcaps · milanzongor/unihack_2020 · kubantjan/fast-form	2019
15	ExquisiteNetV2	0.29	99.71	518230	–	–	–	A Novel lightweight Convolutional Neural Network, ExquisiteNetV2	shyhyawJou/ExquisiteNetV2	2021
15	WaveMix-128/7	0.29	–	–	–	–	–	WaveMix: Resource-efficient Token Mixing for Images	pranavphoenix/WaveMix	2022
18	Fractional MP	0.3	–	–	–	–	–	Fractional Max-Pooling	facebookresearch/SparseConvNet · laplacetw/vgg-like-cifar10 · VladimirGol/KerasFractionalMaxPooling · +2	2014
18	Tree+Max-Avg pooling	0.3	–	–	–	–	–	Generalizing Pooling Functions in Convolutional Neural Networks: Mixed, Gated, and Tree	cypw/DPNs · BeanGreen247/Python-AI-Arts	2015
18	CMsC	0.3	–	–	–	–	–	Competitive Multi-scale Convolution		2015
21	EXACT (M3-CNN)	0.33	–	–	–	–	–	EXACT: How to Train Your Accuracy	tinkoff-ai/exact · ivan-chai/exact	2022
22	Second Order Neural Ordinary Differential Equation	0.37	99.63	–	–	–	–	On Second Order Behaviour in Augmented Neural ODEs	a-norcliffe/sonode	2020
22	Augmented Neural Ordinary Differential Equation	0.37	99.63	–	–	–	–	Augmented Neural ODEs	EmilienDupont/augmented-neural-odes · mitmath/18S096SciML · locuslab/monotone_op_net · +3	2019
24	DSN	0.4	–	–	–	–	–	Deeply-Supervised Nets	ellisdg/3DUnetCNN	2014
24	CKN	0.4	–	–	–	–	–	Convolutional Kernel Networks		2014
24	C-SVDDNet	0.4	–	–	–	–	–	Unsupervised Feature Learning with C-SVDDNet		2014
24	HOPE	0.4	–	–	–	–	–	Hybrid Orthogonal Projection and Estimation (HOPE): A New Framework to Probe and Learn Neural Networks		2015
24	FLSCNN	0.4	–	–	–	–	–	Enhanced Image Classification With a Fast-Learning Shallow Convolutional Neural Network		2015
24	MIM	0.4	–	–	–	–	–	On the Importance of Normalisation Layers in Deep Learning with Piecewise Linear Activation Units		2015
24	Fitnet-LSUV-SVM	0.4	–	–	–	–	–	All you need is a good init	ducha-aiki/LSUVinit · ducha-aiki/LSUV-keras · ducha-aiki/LSUV-pytorch · +8	2015
31	TAAF-CNN	0.48%	99.52%	421642	0.0188	35	–	Evaluating the Performance of TAAF for image classification models	bryn-gnolbs/TAAF-for-Image-Classification	2025
32	Neural Architecture Search (NAS)-enabled Convolutional Neural Network (CNN)	0.5	99.5	1882602	–	–	–
32	Maxout Networks	0.5	–	–	–	–	–	Maxout Networks	MaximeVandegar/Papers-in-100-Lines-of-Code · philipperemy/tensorflow-maxout · mavenlin/cuda-convnet · +4	2013
32	NiN	0.5	–	–	–	–	–	Network In Network	MaximeVandegar/Papers-in-100-Lines-of-Code · nagadomi/kaggle-cifar10-torch7 · yuxiangalvin/DeepLOB-Model-Implementation-Project · +14	2013
32	ReNet	0.5	–	–	–	–	–	ReNet: A Recurrent Neural Network Based Alternative to Convolutional Networks	fvisin/reseg · brian-moser/dartsrenet · NisTa24/ReNet-Implementation · +1	2015
32	DCNN+GFE	0.5	–	–	–	–	–	Deep Convolutional Neural Networks as Generic Feature Extractors		2017
32	VDN	0.5	–	–	–	–	–	Training Very Deep Networks	LiyuanLucasLiu/LM-LSTM-CRF · yoonkim/lstm-char-cnn · flukeskywalker/highway-networks	2015
38	NeuPDE	0.51	–	–	–	–	–	NeuPDE: Neural Network Based Ordinary and Partial Differential Equations for Modeling Time-Dependent Data		2019
39	Simple CNN with BaikalCMA loss	0.53	–	–	–	–	–	Improved Training Speed, Accuracy, and Data Utilization Through Loss Function Optimization	sgonzalez/SwiftGenetics · sgonzalez/SwiftCMA	2019
40	SEER (RegNet10B)	0.58	99.42	–	–	–	–	Vision Models Are More Robust And Fair When Pretrained On Uncurated Images Without Supervision	facebookresearch/vissl	2022
41	Convolutional Tsetlin Machine	0.6	99.4	–	–	–	–	The Convolutional Tsetlin Machine	cair/TsetlinMachine · cair/pyTsetlinMachine · cair/convolutional-tsetlin-machine-tutorial · +6	2019
41	PCANet	0.6	–	–	–	–	–	PCANet: A Simple Deep Learning Baseline for Image Classification?	Ldpe2G/PCANet · lucasleesw/PCANet_python_ver	2014
41	DiffPrune (LeNet5)	0.6	–	–	–	–	–	DiffPrune: Neural Network Pruning with Deterministic Approximate Binary Gates and $L_0$ Regularization	YanivShu/diffprune_public	2020
44	Deep Fried Convnets	0.7	–	–	–	–	–	Deep Fried Convnets	v0lta/wavelet-network-compression	2014
45	Sparse Activity and Sparse Connectivity in Supervised Learning	0.8	–	–	–	–	–	Sparse Activity and Sparse Connectivity in Supervised Learning		2016
45	Explaining and Harnessing Adversarial Examples	0.8	–	–	–	–	–	Explaining and Harnessing Adversarial Examples	cleverhans-lab/cleverhans · openai/cleverhans · tensorflow/cleverhans · +56	2014
47	BinaryConnect	1.0	–	–	–	–	–	BinaryConnect: Training Deep Neural Networks with binary weights during propagations	tensorpack/tensorpack · MatthieuCourbariaux/BinaryConnect · ryuz/BinaryBrain · +2	2015
48	Convolutional PMM (Parametric Matrix Model)	1.01	98.99	129416	–	–	–	Parametric Matrix Models		2024
49	LeNet 300-100 (Sparse Momentum)	1.26	–	–	–	–	–	Sparse Networks from Scratch: Faster Training without Losing Performance	TimDettmers/sparse_learning · google-research/rigl	2019
50	Convolutional Clustering	1.4	–	–	–	–	–	Convolutional Clustering for Unsupervised Learning		2015
51	CNN Model by Som	1.41	98.59	–	–	–	–	Convolutional Sequence to Sequence Learning	facebookresearch/fairseq · facebookresearch/ParlAI · OpenNMT/OpenNMT-py · +34	2017
52	Weighted Tsetlin Machine	1.5	98.5	–	–	–	–	The Weighted Tsetlin Machine: Compressed Representations with Weighted Clauses	cair/pyTsetlinMachine · adrianphoulady/weighted-tsetlin-machine-cpp	2019
53	MLP (ideal number of groups)	1.67	–	–	–	–	–	On the Ideal Number of Groups for Isometric Gradient Propagation		2023
54	Perceptron with a tensor train layer	1.8	98.2	–	–	–	–	Tensorizing Neural Networks	timgaripov/TensorNet-TF · Bihaqo/TensorNet · philip-bl/tensorizing_neural_networks-novikov_2015 · +1	2015
54	ANODE	1.8	98.2	–	–	–	–	Augmented Neural ODEs	EmilienDupont/augmented-neural-odes · mitmath/18S096SciML · locuslab/monotone_op_net · +3	2019
54	Tsetlin Machine	1.8	98.2	–	–	–	–	The Tsetlin Machine - A Game Theoretic Bandit Driven Approach to Optimal Pattern Recognition with Propositional Logic	cair/TsetlinMachine · cair/pyTsetlinMachine · cair/fast-tsetlin-machine-with-mnist-demo · +13	2018
57	GECCO	1.96	98.04	–	–	–	–	A Single Graph Convolution Is All You Need: Efficient Grayscale Image Classification	geccoproject/gecco	2024
58	PMM (Parametric Matrix Model)	2.62	97.38	4990	–	–	–	Parametric Matrix Models		2024
59	DNN-5 (Trainable Activations)	2.8	97.2	575051	–	–	–	Trainable Activations for Image Classification	Pe4enIks/TrainableActivation	2023
60	DNN-3 (Trainable Activations)	3.0	97.0	386719	–	–	–	Trainable Activations for Image Classification	Pe4enIks/TrainableActivation	2023
61	DNN-2 (Trainable Activations)	3.6	96.4	311651	–	–	–	Trainable Activations for Image Classification	Pe4enIks/TrainableActivation	2023
62	Zhao et al. (2015) (auto-encoder)	4.76	–	–	–	–	–	Stacked What-Where Auto-encoders	isaacgerg/keras_odds_and_ends · zhangqinghao0811/unpool	2015
63	ProjectionNet	5.0	95.0	–	–	–	–	ProjectionNet: Learning Efficient On-Device Deep Networks Using Neural Projections		2017
64	µ2Net (ViT-L/16)	–	99.75	–	–	–	–	An Evolutionary Approach to Dynamic Introduction of Tasks in Large-scale Multitask Learning Systems	google-research/google-research	2022
65	MobileNet_XnODR	–	99.68	–	–	–	–	XnODR and XnIDR: Two Accurate and Fast Fully Connected Layers For Convolutional Neural Networks	jiansfoggy/CODE-SHOW	2021
66	ResNet-9	–	99.68	–	–	–	–	CNN Filter DB: An Empirical Investigation of Trained Convolutional Filters	paulgavrikov/cnn-filter-db	2022
67	LR-Net	–	99.47	–	–	–	–	LR-Net: A Block-based Convolutional Neural Network for Low-Resolution Image Classification	AshkanGanj/Block-Based-ImageClassification-Architecture	2022
68	CNN+ Wilson-Cowan model RNN	–	99.31	–	–	–	–	Learning in Wilson-Cowan model for metapopulation	raffaelemarino/learning_in_wilsoncowan	2024
69	FastSNN (CNN)	–	99.3	–	–	–	–	Robust and accelerated single-spike spiking neural network training with applicability to challenging temporal tasks	webstorms/fastsnn	2022
70	rKAN	–	99.293	–	–	–	–	rKAN: Rational Kolmogorov-Arnold Networks	alirezaafzalaghaei/rkan	2024
71	CNN-5 Layer	–	99.27	–	–	–	–	Robust Training in High Dimensions via Block Coordinate Geometric Median Descent	anishacharya/BGMD · anishacharya/Optimization-Mavericks	2021
72	fKAN	–	99.228	–	–	–	–	fKAN: Fractional Kolmogorov-Arnold Networks with trainable Jacobi basis functions	alirezaafzalaghaei/fKAN	2024
73	StiDi-BP in R-CSNN	–	99.2	–	–	–	–	Spike time displacement based error backpropagation in convolutional spiking neural networks		2021
74	Wilson-Cowan model RNN	–	98.13	–	–	–	–	Learning in Wilson-Cowan model for metapopulation	raffaelemarino/learning_in_wilsoncowan	2024
75	Hypervector Tsetlin Machine	–	98.13	–	–	–	–	Exploring Effects of Hyperdimensional Vectors for Tsetlin Machines		2024
76	ViT-Mini_D9	–	98.03	1208586	–	–	–
77	FastSNN (MLP)	–	97.91	–	–	–	–	Robust and accelerated single-spike spiking neural network training with applicability to challenging temporal tasks	webstorms/fastsnn	2022
78	Binarized MLP with on-chip spiking backpropagation (on Loihi)	–	96.2	–	–	–	–	The Backpropagation Algorithm Implemented on Spiking Neuromorphic Hardware	lanl/spikingBackprop	2021
79	SNNL-5	–	95.5	–	–	–	–	Improving k-Means Clustering Performance with Disentangled Internal Representations	afagarap/pt-snnl	2020
80	pFedBreD_ns_mg	–	92.47	–	–	–	–	Personalized Federated Learning with Hidden Information on Personalized Prior		2022
81	DGMMC-S	–	–	–	–	–	70	Performance of Gaussian Mixture Model Classifiers on Embedded Feature Spaces	cvmlmu/dgmmc	2024
82	HiRo 자동 추출	–	99.46	–	–	–	–	HiRo: A Compact Four-Directional Hierarchical Reservoir Token-Mixer for Efficient Image Classification		2026
83	On the Role of Preprocessing and Memrist 자동 추출	–	95.89	–	–	–	–	On the Role of Preprocessing and Memristor Dynamics in Reservoir Computing for Image Classification		2026
84	Silicon Aware Neural Networks 자동 추출	–	97	–	–	–	–	Silicon Aware Neural Networks		2026
85	Photonic AI 자동 추출	–	91.2	–	–	–	–	Photonic AI: A Hybrid Diffractive Holographic Neural System for Passive Optical Real-Time Image Classification		2026
86	Machine vision with small numbers of det 자동 추출	–	97	–	–	–	–	Machine vision with small numbers of detected photons per inference		2026
87	Growing 자동 추출	–	99.44	–	–	–	–	Growing Networks with Autonomous Pruning		2026
88	QCNN 자동 추출	–	98.7	–	–	–	–	Beyond Barren Plateaus: A Scalable Quantum Convolutional Architecture for High-Fidelity Image Classification		2026
89	Primitive-Driven Acceleration of Hyperdi 자동 추출	–	95.67	–	–	–	–	Primitive-Driven Acceleration of Hyperdimensional Computing for Real-Time Image Classification		2026
90	Enhancing Small Dataset Classification U 자동 추출	–	95	–	–	–	–	Enhancing Small Dataset Classification Using Projected Quantum Kernels with Convolutional Neural Networks		2026
91	Multi-Scale Visual Prompting for Lightwe 자동 추출	–	–	0.02	–	–	–	Multi-Scale Visual Prompting for Lightweight Small-Image Classification		2025
92	A Methodology for Transparent Logic-Base 자동 추출	–	98.5	–	–	–	–	A Methodology for Transparent Logic-Based Classification Using a Multi-Task Convolutional Tsetlin Machine		2025