berksafran / train-faces

UI for training faces for Microsoft Cognitive Services Face API

Cognitive Services Face API Training

The goal of this project is to create a dashboard for users to manage and test their Cognitive Services Face API, given that they have a Face API Key

Features

• Manages your Person, Person Group, and Training of your Face API
• Locally Storing your Face API Key
• Nicely formated User Data for the Person entities
• Quickly submit Face entities using the webcam

Getting Started

You can go to the Live App.

Or if you want to Spin up your own instance of this app, you will need node, npm or yarn:

1. clone this repository
2. run npm install (this will take a little while)
3. run yarn start OR yarn run compile and serve the assets in the dist directory
4. go to http://localhost:3000

---

Technical Details

Key Application Dependencies

Backend This app is now serverless :)

• express (Server)
• project-oxford (Face API Client)

Frontend

• react (View Layer)
• redux (App State Management)
• react-router (Router for React)
• [redux-thunk]((https://www.npmjs.com/package/react-router) (Redux Async Management)
• antd (React Component Library)
• react-json-tree (Render Json Tree)
• react-ace (AceEditor for react)
• lodash.get/set (Access deeply nested values)

Development Tools

• webpack (Asset/Package Bundling)
• babel (Transpiles ES7 to ES5)
• webpack-hot-middleware (Hot Module Replace)
• plop (Code Generator)
• eslint (Code Style Linter)

Getting Started with Development

1. clone this repository
2. run yarn (might take a while)
3. run yarn start
4. go to localhost:3000

To Contribute

Please refer to the CONTRIBUTING.md

TODO

• Async Flow in UI
• Auto Refresh on FaceAPIKey change
• Mobile Friendly
• Fix User Data for Face Creation
• Add User Data for Person Group Creation
• Convert ternary operators to _.get for readability
• Build dist on server
• Helpful Error Messages