官网: https://wgai.dromara.org/#/
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;
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
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
@github https://github.com/YeyuchenBa/wgai
Online training and annotation content have been launched and open source support is available for deployment on domestically produced Windows/Linux servers
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
Regulatory Redis jvm、 The CPU of the server is particularly important
Independently adding trained models, training models and recognition are separated to avoid memory consumption
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
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
Support third-party interface transmission, recognize continuously after enabling without interrupting sub threads. CPU running requires a certain amount of memory
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)
Support text language models such as Exel and TXT; Scenario specific training
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
##为大家低成本部署AI应用添砖加瓦！
Technical questions will be answered within the community, along with various installation environment configurations
