# captcha-dataset

**Repository Path**: JinMana/captcha-dataset

## Basic Information

- **Project Name**: captcha-dataset
- **Description**: Captcha-dataset: one captcha dataset for tensorflow/torch7/keras/caffe2 一个验证码的数据集
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 1
- **Created**: 2020-03-24
- **Last Updated**: 2020-12-19

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Captcha-dataset: one captcha dataset for tensorflow/torch7/keras/caffe2 一个验证码的数据集



Sample:

![](0a1d2800-33ec-424f-8a3a-8e5b1ee2726b.png)

![](0a4b208f-3100-44a6-b530-dbe3670a480c.png)



Data:

```
tar -ezf yzm.tar.gz
tar -ezf yzm1.tar.gz
mv yzm1/* yzm/
```

(Because of github's 100MB limitation)



**9272 in total, made by human labeling, accuracy > 98%**

Label file is `yzm_labels.txt`



Useful Res:

Related repo [decaptcha](https://github.com/fateleak/decaptcha)

tensorflow共享群(625706505)：

![](tensorflow-CNN-share.jpeg)