# WGAI
**Repository Path**: normalcoder/wgai
## Basic Information
- **Project Name**: WGAI
- **Description**: 开箱即用的JAVAAI在线训练识别平台&OCR平台AI合集包含旦不仅限于(车牌识别、安全帽识别、抽烟识别、常用类物识别等) 图片和视频识别 可自主
融合了AI图像识别opencv、yolo、ocr、esayAI内核识别;AI智能客服、AI语言模型、 无任何第三方API接口可定制化自主离线化部署并自主化行业化使用
避免占用内存、GPU消耗训练与识别分开使用;
- **Primary Language**: Java
- **License**: Apache-2.0
- **Default Branch**: master
- **Homepage**: http://116.198.227.105/
- **GVP Project**: No
## Statistics
- **Stars**: 0
- **Forks**: 536
- **Created**: 2025-02-12
- **Last Updated**: 2025-02-12
## Categories & Tags
**Categories**: Uncategorized
**Tags**: None
## README
**官网: [http://116.198.227.105/#/](http://116.198.227.105/#/)**
[](https://discord.gg/Fb6M73htGr)
[](https://www.reddit.com/r/hertzbeat/)
[](https://www.bestpractices.dev/projects/8139)
[](https://hub.docker.com/r/apache/hertzbeat)
[](https://gitpod.io/#https://github.com/apache/hertzbeat)
## 🎡 introduce
> The out of the box JAVA AI platform integrates AI image recognition with OpenCV, YOLO, and eSayAI kernel recognition; AI intelligent customer service, AI language model, customizable and independent offline deployment, and industry-specific use
Avoid using memory and GPU separately for training and recognition; Support YOLOv3, YOLOv5, YOLOv8 models and text language models such as Excel and TXT;
### Test Address&Account
> [ http://116.198.227.105:8888](http://116.198.227.105:8888/)
> wgai wgai@2024
>
> Note: The server purchased for 50 yuan crashes if it runs away. It is recommended not to use it violently
# Get started directly
## Download to use
> springboot+vue+mysqlSupport for domestic database configuration content will not be elaborated in detail. If Java is weak, it is recommended to learn Java first and then use the front-end in the VUE branch!!!
> [@gitee https://gitee.com/wggh_admin/wgai](https://gitee.com/wggh_admin/wgai)
>[@github https://github.com/YeyuchenBa/wgai](https://github.com/YeyuchenBa/wgai)
# New feature screenshot and feature introduction configuration
### Online training and annotation content have been launched and open source support is available for deployment on domestically produced Windows/Linux servers
#### Model List
#### Online annotation
#### training results
#### Training log
### New feature demonstration video
>[@Online training demonstration video address：https://www.bilibili.com/video/BV1EJwheEEYq/?vd_source=73d8a30a253a95bdb8b89a5fec80d9b9](https://www.bilibili.com/video/BV1EJwheEEYq/?vd_source=73d8a30a253a95bdb8b89a5fec80d9b9)
>[@Online training demonstration video address：https://www.bilibili.com/video/BV1fJwhe7E1G/?vd_source=73d8a30a253a95bdb8b89a5fec80d9b9](https://www.bilibili.com/video/BV1fJwhe7E1G/?vd_source=73d8a30a253a95bdb8b89a5fec80d9b9)
## Localized deployment of voice recognition and addition of hot word configuration support for Windows/Linux localization server deployment
## Currently supports voice recognition, but since I don't have HTTPS, I have added an additional static audio recognition feature
## Currently supports blue, green, yellow, white, and vehicle license plate recognition
## Currently supports OCR 95% high-precision text extraction and recognition
# Currently, feature screenshots and feature descriptions are supported for configuration
## - Autonomous image and video recognition
### home page
>Regulatory Redis jvm、 The CPU of the server is particularly important
### model base
>Independently adding trained models, training models and recognition are separated to avoid memory consumption
### Model library binding
>Support image upload, image address, video address rtsp、rtmp、flv、fmp4、 Does not support static MP4 playback, but supports recognition; Because the player control does not allow static file playback
### Image recognition
>Support third-party interface transmission with recognition time generally within 1 second, except for special and complex issues. The current unit of time is in milliseconds
### Video recognition
>Support third-party interface transmission, recognize continuously after enabling without interrupting sub threads. CPU running requires a certain amount of memory
### Third party access
>Third party access
> The currently supported item recognition settings include Chinese translation and border color (continuously updated). If you have any specific model requirements, please contact me 1552138571@qq.com Or qq: 1552138571)
## - ChatGpt
### chatGPT
> Support text language models such as Exel and TXT; Scenario specific training
### Image model
### Training results
### Training tasks
## - Lightweight intelligent chat
> This lightweight kernel mainly uses the easyui kernel [@ easyAi Lightweight Kernel Address]（ https://gitee.com/dromara/easyAi ）They have their own content for training, with low memory consumption and short training time. They have high efficiency in recognizing specific content
> Due to limited time, students who have not yet fully connected can continue to connect independently
### Basic classification
### Semantic classification
### Sentence classification
### Keywords:
### Intelligent dialogue
### QA Q&A
### Semantic model training
##为大家低成本部署AI应用添砖加瓦！
## We will continue to optimize video and image recognition content in the future
### Technical questions will be answered within the community, along with various installation environment configurations
### WeChat communication group
## Add WeChat technical communication group (currently only WeChat communication group)
## 🛡️ License
[`Apache License, Version 2.0`](https://www.apache.org/licenses/LICENSE-2.0.html)