https://github.com/tansen87/handwrittenformularecognition

Handwritten Formula Recognition | 手写公式识别
https://github.com/tansen87/handwrittenformularecognition

handwriting-recognition handwritten-digit-recognition math-formula-recognition pytorch resnet

Last synced: 10 months ago
JSON representation

Handwritten Formula Recognition | 手写公式识别

• Host: GitHub
• URL: https://github.com/tansen87/handwrittenformularecognition
• Owner: tansen87
• License: gpl-3.0
• Created: 2022-05-19T11:45:59.000Z (about 4 years ago)
• Default Branch: main
• Last Pushed: 2024-12-25T04:03:24.000Z (over 1 year ago)
• Last Synced: 2025-07-09T20:51:45.689Z (about 1 year ago)
• Topics: handwriting-recognition, handwritten-digit-recognition, math-formula-recognition, pytorch, resnet
• Language: Python
• Homepage:
• Size: 38.9 MB
• Stars: 18
• Watchers: 1
• Forks: 3
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# 基于ResNet的手写公式识别

> Notice: python版本使用 ^3.9

![image](https://github.com/tansen87/HandwrittenFormulaRecognition/assets/98570790/13bf3de4-c238-4aa4-b6e2-23e6808ee08e)

### 运行项目

1. 安装uv

   ```bash

   pip install uv

   ```

2. 使用uv创建虚拟环境

   ```bash

   uv venv

   ```

3. 激活创建的虚拟环境

   ```bash

   .venv\Scripts\activate

   ```

4. 安装依赖

   ```bash

   uv pip install -r requirements.txt

   ```

5. 运行项目

   ```bash

   uv run main.py

   ```

#### video demo

[bilibili](https://www.bilibili.com/video/BV1Qa411f7dB/?spm_id_from=333.999.0.0&vd_source=5ee5270944c6e7a459e1311330bf455c)

#### tips

* 目前只训练了0, 1, 2, 3, 4, 5, 6, 7, 8, 9 , +, -, =,*, /, (, )符号

* 如果你想自己训练，在[kaggle](https://www.kaggle.com/xainano/handwrittenmathsymbols)下载数据集，然后`train/train.py`进行训练