Metric: Accuracy (higher is better)
| # | Model↕ | Accuracy▼ | Extra Data | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | WaveMixLite-112/16 | 95.96 | No | WaveMix: A Resource-efficient Neural Network for... | 2022-05-28 | Code |
| 2 | MIN-ENT + VGG-5 | 95.933 | No | Regularizing cross entropy loss via minimum entr... | 2025-01-23 | Code |
| 3 | MIX-ENT + VGG-5 | 95.928 | No | Regularizing cross entropy loss via minimum entr... | 2025-01-23 | Code |
| 4 | VGG-5(Spinal FC) | 95.88 | No | SpinalNet: Deep Neural Network with Gradual Input | 2020-07-07 | Code |
| 5 | VGG-5 | 95.86 | No | ProgressiveSpinalNet architecture for FC layers | 2021-03-21 | Code |
| 6 | VGG-5 | 95.86 | No | ProgressiveSpinalNet architecture for FC layers | 2021-03-21 | Code |
| 7 | TextCaps | 95.39 | No | TextCaps : Handwritten Character Recognition wit... | 2019-04-17 | Code |
| 8 | µ2Net+ (ViT-L/16) | 95.03 | No | A Continual Development Methodology for Large-sc... | 2022-09-15 | Code |
| 9 | OptConv+Log+Perc | 93.65 | No | - | - | - |
| 10 | DWT-DCT + SVM | 89.51 | No | - | - | - |
| 11 | OPIUM Classifier | 85.27 | No | EMNIST: an extension of MNIST to handwritten let... | 2017-02-17 | Code |
| 12 | Linear Classifier | 55.78 | No | EMNIST: an extension of MNIST to handwritten let... | 2017-02-17 | Code |