可用于甲骨文字符的训练与识别,支持上传图片自动检测横竖排、智能分割并实时识别。基于ResNet18模型,包含353个字符库,采用PyTorch训练,提供Flask Web演示。【此简介由AI生成】
| Files | Last commit | Last update |
|---|
| init new structure | 10 months ago |
| increment training for m2, and according predict web app | 10 months ago |
| increment training for m2, and according predict web app | 10 months ago |
| add yolo detect model | 10 months ago |
| add arcface loss training and web test app | 10 months ago |
| add arcface loss training and web test app | 10 months ago |
| update: 更新文件 README.md Signed-off-by: oasisszli <oasisszli@noreply.gitcode.com> | 10 months ago |
| add yolo detect model | 10 months ago |
| more ignore | 10 months ago |
| more ignore | 10 months ago |
| more ignore | 10 months ago |
| more ignore | 10 months ago |
| more ignore | 10 months ago |
| more ignore | 10 months ago |
| more ignore | 10 months ago |
| now we trained 353 chars,实现了竖排切割拓片简单算法,并且测试网站更新支持拓片分割 | 10 months ago |
| support 人工标记拓片 | 10 months ago |
| init new structure | 10 months ago |
| init new structure | 10 months ago |
| init new structure | 10 months ago |
| init new structure | 10 months ago |
| init new structure | 10 months ago |
| 支持竖排横排单行拓片切字简单算法 | 10 months ago |
| init new structure | 10 months ago |
| init new structure | 10 months ago |
| init new structure | 10 months ago |
| init new structure | 10 months ago |
| init new structure | 10 months ago |
| init new structure | 10 months ago |
| init new structure | 10 months ago |
| init new structure | 10 months ago |
| support 人工标记拓片 | 10 months ago |
甲骨文识别系统
基于深度学习的甲骨文字符训练与识别系统。
训练数据
- 字符库: 353个甲骨文字符
- 模型: ResNet18
- 数据: 210data目录
Web演示



cd web && python app.py
功能:
流程:
- 上传图片
- 选择"拓片"模式
- 查看分割
- 点击识别
识别功能
- 模型: ResNet18
- 字符: 353个
- 输出: Top-10预测
技术: PyTorch训练、ResNet18模型、Flask演示、210data字库
特点: 自动分割、智能识别、实时演示
可用于甲骨文字符的训练与识别,支持上传图片自动检测横竖排、智能分割并实时识别。基于ResNet18模型,包含353个字符库,采用PyTorch训练,提供Flask Web演示。【此简介由AI生成】
Customize your domain