ML Research Wiki / Benchmarks / Image Classification / Kuzushiji-MNIST

Kuzushiji-MNIST

Image Classification Benchmark

Performance Over Time

📊 Showing 14 results | 📏 Metric: Accuracy

Top Performing Models

Rank	Model	Paper	Accuracy	Date	Code
1	VGG-5 (Spinal FC)	SpinalNet: Deep Neural Network with Gradual Input	99.15	2020-07-07	📦 dipuk0506/SpinalNet 📦 dipuk0506/uq 📦 Mechachleopteryx/SpinalNet
2	CAMNet3	Context-Aware Multipath Networks	99.05	2019-07-26	-
3	VGG8B(2x) + LocalLearning + CO	Training Neural Networks with Local Error Signals	99.01	2019-01-20	📦 anokland/local-loss 📦 ai-tech-research-lab/nitro-d
4	Resnet-152	A Comprehensive Study of ImageNet Pre-Training for Historical Document Image Analysis	98.79	2019-05-22	-
5	R-ExplaiNet-26	Learning local discrete features in explainable-by-design convolutional neural networks	98.78	2024-10-31	📦 pikaplan/LearnExplaiNet
6	ResNet-14	CNN Filter DB: An Empirical Investigation of Trained Convolutional Filters	98.75	2022-03-29	📦 paulgavrikov/cnn-filter-db
7	PreActResNet-18 + Input Mixup	mixup: Beyond Empirical Risk Minimization	98.41	2017-10-25	📦 rwightman/pytorch-image-models 📦 pytorch/vision 📦 PaddlePaddle/PaddleDetection
8	PreActResNet-18	Identity Mappings in Deep Residual Networks	97.82	2016-03-16	📦 tensorflow/models 📦 tensorflow/models 📦 tensorflow/models
9	Convolutional Tsetlin Machine	The Convolutional Tsetlin Machine	96.30	2019-05-23	📦 cair/TsetlinMachine 📦 cair/pyTsetlinMachine 📦 cair/convolutional-tsetlin-machine-tutorial
10	KerCNN	KerCNNs: biologically inspired lateral connections for classification of corrupted images	93.13	2019-10-18	-

All Papers (14)

SpinalNet: Deep Neural Network with Gradual Input

2020

VGG-5 (Spinal FC)

dipuk0506/SpinalNet dipuk0506/uq Mechachleopteryx/SpinalNet

Context-Aware Multipath Networks

2019

CAMNet3

Training Neural Networks with Local Error Signals

2019

VGG8B(2x) + LocalLearning + CO

anokland/local-loss ai-tech-research-lab/nitro-d

A Comprehensive Study of ImageNet Pre-Training for Historical Document Image Analysis

2019

Resnet-152

Learning local discrete features in explainable-by-design convolutional neural networks

2024

R-ExplaiNet-26

pikaplan/LearnExplaiNet

CNN Filter DB: An Empirical Investigation of Trained Convolutional Filters

2022

ResNet-14

paulgavrikov/cnn-filter-db

mixup: Beyond Empirical Risk Minimization

2017

PreActResNet-18 + Input Mixup

rwightman/pytorch-image-models pytorch/vision

Identity Mappings in Deep Residual Networks

2016

PreActResNet-18

tensorflow/models tensorflow/models

The Convolutional Tsetlin Machine

2019

Convolutional Tsetlin Machine

cair/TsetlinMachine cair/pyTsetlinMachine

KerCNNs: biologically inspired lateral connections for classification of corrupted images

2019

KerCNN

Multi-Complementary and Unlabeled Learning for Arbitrary Losses and Models

2020

linear/flexible model

Multi-Complementary and Unlabeled Learning for Arbitrary Losses and Models

2020

FWD

Complementary-Label Learning for Arbitrary Losses and Models

2018

Complementary-Label Learning

takashiishida/comp

Deep Learning for Classical Japanese Literature

2018

ResNet18 + VGG Ensemble

rois-codh/kmnist christianversloot/extra_keras_datasets

Kuzushiji-MNIST

Performance Over Time

Edit Benchmark Results

Edit Result

Top Performing Models

All Papers (14)

SpinalNet: Deep Neural Network with Gradual Input

Context-Aware Multipath Networks

Training Neural Networks with Local Error Signals

A Comprehensive Study of ImageNet Pre-Training for Historical Document Image Analysis

Learning local discrete features in explainable-by-design convolutional neural networks

CNN Filter DB: An Empirical Investigation of Trained Convolutional Filters

mixup: Beyond Empirical Risk Minimization

Identity Mappings in Deep Residual Networks

The Convolutional Tsetlin Machine

KerCNNs: biologically inspired lateral connections for classification of corrupted images

Multi-Complementary and Unlabeled Learning for Arbitrary Losses and Models

Multi-Complementary and Unlabeled Learning for Arbitrary Losses and Models

Complementary-Label Learning for Arbitrary Losses and Models

Deep Learning for Classical Japanese Literature

Model	Paper	Accuracy	Date
VGG-5 (Spinal FC)	SpinalNet: Deep Neural Network with Gradual Input	99.15	2020-07-07
CAMNet3	Context-Aware Multipath Networks	99.05	2019-07-26
VGG8B(2x) + LocalLearning + CO	Training Neural Networks with Local Error Signals	99.01	2019-01-20
Resnet-152	A Comprehensive Study of ImageNet Pre-Training fo…	98.79	2019-05-22
R-ExplaiNet-26	Learning local discrete features in explainable-b…	98.78	2024-10-31
ResNet-14	CNN Filter DB: An Empirical Investigation of Trai…	98.75	2022-03-29
PreActResNet-18 + Input Mixup	mixup: Beyond Empirical Risk Minimization	98.41	2017-10-25
PreActResNet-18	Identity Mappings in Deep Residual Networks	97.82	2016-03-16
Convolutional Tsetlin Machine	The Convolutional Tsetlin Machine	96.30	2019-05-23
KerCNN	KerCNNs: biologically inspired lateral connection…	93.13	2019-10-18
linear/flexible model	Multi-Complementary and Unlabeled Learning for Ar…	79.90	2020-01-13
FWD	Multi-Complementary and Unlabeled Learning for Ar…	79.50	2020-01-13
Complementary-Label Learning	Complementary-Label Learning for Arbitrary Losses…	67.10	2018-10-10
ResNet18 + VGG Ensemble	Deep Learning for Classical Japanese Literature	1.10	2018-12-03