# Image Classification on Kuzushiji-MNIST (KMNIST) (test)

Source: [Image Classification on Kuzushiji-MNIST (KMNIST) (test)](<https://www.sota2.com/research/sota/image-classification-on-kuzushiji-mnist-kmnist-test>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T06:06:14.314Z

Evaluation of Image Classification performance on the test split of the Kuzushiji-MNIST (KMNIST) dataset.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Image Classification |
| Dataset (including any stated split) | Kuzushiji-MNIST (KMNIST) (test) |
| Metric | Test Accuracy (Avg) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 5 of 5 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Test Accuracy (Avg)](<https://www.sota2.com/research/sota/image-classification-on-kuzushiji-mnist-kmnist-test.md?metric=Test%20Accuracy%20(Avg)>): Higher is better; unit: PERCENT.

- [Test Accuracy (Best)](<https://www.sota2.com/research/sota/image-classification-on-kuzushiji-mnist-kmnist-test.md?metric=Test%20Accuracy%20(Best)>): Higher is better; unit: PERCENT.

- [Error Reduction (Best)](<https://www.sota2.com/research/sota/image-classification-on-kuzushiji-mnist-kmnist-test.md?metric=Error%20Reduction%20(Best)>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Test Accuracy (Avg) | Result date | Report paper |
| --- | --- | --- | --- |
| VGG-5 (Spinal FC) | 99.05 | Not recorded | 68fa74271e209131cea600e4 |
| VGG-5 | 98.76 | Not recorded | 68fa74271e209131cea600e4 |
| CNN (Spinal FC) | 86.48 | Not recorded | 68fa74271e209131cea600e4 |
| CNN (Spinal FC) | 85.98 | Not recorded | 68fa74271e209131cea600e4 |
| CNN | 83.24 | Not recorded | 68fa74271e209131cea600e4 |

### VGG-5 (Spinal FC)

Test Accuracy (Avg): 99.05. Result ID: 6993337c3a6c314849146639.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74271e209131cea600e4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74271e209131cea600e4.md>)

[Implementation](<https://github.com/dipuk0506/SpinalNet>)

Recorded evaluation configuration: {"Size of Fully Connected Layer":"4HL, 128 Neurons Per Layer","Epoch":200,"Parameters":"3.630M"}. This record does not assert independent reproduction.

### VGG-5

Test Accuracy (Avg): 98.76. Result ID: 6993337c3a6c314849146638.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74271e209131cea600e4.md>)

Recorded evaluation configuration: {"Size of Fully Connected Layer":"1HL, 512 Neurons","Epoch":200,"Parameters":"3.646M"}. This record does not assert independent reproduction.

### CNN (Spinal FC)

Test Accuracy (Avg): 86.48. Result ID: 6993337c3a6c314849146637.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74271e209131cea600e4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74271e209131cea600e4.md>)

[Implementation](<https://github.com/dipuk0506/SpinalNet>)

Recorded evaluation configuration: {"Size of Fully Connected Layer":"6HL, 10 Neurons Per Layer","Epoch":8,"Parameters":"16.05k"}. This record does not assert independent reproduction.

### CNN (Spinal FC)

Test Accuracy (Avg): 85.98. Result ID: 6993337c3a6c314849146636.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74271e209131cea600e4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74271e209131cea600e4.md>)

[Implementation](<https://github.com/dipuk0506/SpinalNet>)

Recorded evaluation configuration: {"Size of Fully Connected Layer":"6HL, 8 Neurons Per Layer","Epoch":8,"Parameters":"13.82k"}. This record does not assert independent reproduction.

### CNN

Test Accuracy (Avg): 83.24. Result ID: 6993337c3a6c314849146635.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74271e209131cea600e4.md>)

Recorded evaluation configuration: {"Size of Fully Connected Layer":"1HL, 50 Neurons","Epoch":8,"Parameters":"21.84k"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.