Open-source image multi-class classification model - Free implementation of high-accuracy image classification

Image Multi Class Classification Not Evaluated

Developed by autoevaluate

This is an image classification model fine-tuned on the MNIST dataset based on the Swin-Tiny architecture, with an accuracy of 98.33%.

Image Classification

Transformers

Open Source License:Apache-2.0 #High accuracy on MNIST #Swin fine-tuning #Lightweight classification

Downloads 11

Release Time : 12/2/2022

Model Overview

This model is an image classification model fine-tuned on the MNIST handwritten digit dataset based on microsoft/swin-tiny-patch4-window7-224, mainly used to recognize handwritten digits from 0 - 9.

Model Features

High accuracy

Achieved an accuracy of 98.33% on the MNIST test set.

Based on Swin Transformer

Adopts the advanced Swin Transformer architecture, with excellent visual feature extraction capabilities.

Lightweight model

Uses the Tiny version of Swin Transformer, suitable for environments with limited resources.

Model Capabilities

Handwritten digit recognition

Image classification

Digit recognition

Use Cases

Education

Automatic grading of handwritten digits

Automatically recognize students' handwritten digit homework.

Accuracy: 98.33%

Finance

Check digit recognition

Recognize the handwritten amount digits on a check.

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.3743	1.0	422	0.0556	0.9833

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご

Image Multi Class Classification Not Evaluated

Model Overview

Model Features

Model Capabilities

Use Cases

🚀 Image Classification

🚀 Quick Start

📚 Documentation

Model description

Intended uses & limitations

Training and evaluation data

🔧 Technical Details

Training procedure

Training hyperparameters

Training results

Framework versions

📄 License