https://github.com/edgeimpulse/mobile-client
Mobile acquisition and inferencing client
https://github.com/edgeimpulse/mobile-client
Last synced: about 1 year ago
JSON representation
Mobile acquisition and inferencing client
- Host: GitHub
- URL: https://github.com/edgeimpulse/mobile-client
- Owner: edgeimpulse
- License: bsd-3-clause-clear
- Created: 2020-04-02T13:01:19.000Z (about 6 years ago)
- Default Branch: master
- Last Pushed: 2025-02-20T10:58:18.000Z (over 1 year ago)
- Last Synced: 2025-04-02T08:09:05.076Z (about 1 year ago)
- Language: CSS
- Size: 16.9 MB
- Stars: 67
- Watchers: 18
- Forks: 18
- Open Issues: 11
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Edge Impulse Mobile acquisition and inferencing client
## Using this client
You don't need this repository. Just go to the **Devices** tab in Edge Impulse, click **Add a new device** and select **Mobile phone**. This pops up a QR code that you can scan with your phone. That's it!
## Developing with this client
Turns your mobile phone into a fully-supported client for [Edge Impulse](https://www.edgeimpulse.com). You can capture data from the accelerometer and microphone, verify your machine learning model, and run classification without leaving your browser. The easiest way to use this application is to head to the [Edge Impulse documentation: Mobile phone](https://docs.edgeimpulse.com/docs/using-your-mobile-phone).
## Installation
To build:
1. Install dependencies:
```
$ npm install
```
1. Build and serve the application:
```
$ npm run start
```
This automatically watches for changes in the code, and rebuilds the application.
1. You'll need to be connected over HTTPS to access sensors (at least on iPhone), use ngrok to open up your web browser to the world via:
```
$ ngrok http 4820
```
Go to the HTTPS URL that ngrok printed to see the client.
## Attribution
The mobile client was built in conjunction with [Brian Weeteling](https://www.brianweet.com).