AI application in electronic component identification

Demo account: 

-  Electronic components experts: Account:chuyengia@demo.com password:123456

Web AI is designed based on ReactJs technology for FE, NodeJs for BE, built according to the Client-Server model, data saved to MongoDB. The Component Identification section uses pre-train models yolov5 and yolov8. The main purpose of this web AI is to create a data labeling tool inspired by label images

Reason for creating web AI: In the current era, electronic devices are indispensable everywhere in the world. It is the core to convert and control electronic devices, but in some cases it is has been exploited to install a number of malicious components for malicious purposes by some organizations, so using AI to identify strange electronic components of unknown origin and not in the database will be eliminated. Therefore, we have researched and built a toolkit to create a database system to automatically scan and check electronic circuit boards to see if there are harmful components installed.

Web AI is built for 3 audiences: Admin, Electronic components experts, and users
- Admin: Admin has the right to operate and participate in all activities of web AI. An account can be created for an electronic components expert to use the data labeling tool

- Electronic components experts: After an admin creates an account and grants permission to use the data building tool, experts can use the tool to label their own data sets. Data sets created on an expert's account are private only. When an expert invites another expert, that other expert can join the invited expert's data set (inspired by google docs)

- User: Register and log in using the detect tool with 2 model options yolov5 and v8 (similar idea to chatGPT)

Database built in the direction of dynamic data. Because each electronic component, when labeled, not only has a name field and a code field, but also has an origin field, function, ....