Start using Developer Edition

Set up and start using AITRIOS Developer Edition, and then create your own vision AI solution

A quick get started guide

Basic setup

Set up a basic AI project with AITRIOS in just a few steps

Follow these steps under the Related docs section to get started with AITRIOS very quickly using a ready-to-use AI model:

Prepare your edge device
Connect your edge device to Console
Download and deploy a ready-to-use model
Run inference

Related docs

Quick Get Started Guide ❯

End-to-end workflow

Create a project

Start by creating your first project in Portal

Visit Portal to create a project and manage users.

Related docs

Portal User Manual ❯

Connect a device

Connect your edge device to Console

Have one of the following supported edge devices ready:

Sony AIH-IVRW2

Rayprus CSV26

From Portal, subscribe to Console and then launch it from your project dashboard. Connect your edge device to Console by scanning a QR code.

Related docs

Device Setup Guide ❯

Console V2 User Manual ❯

Train AI models

Start developing and training AI models

Now that your device is registered and connected to Console, you can start exploring AI models. Begin with the object detection model and create a Custom Vision project. Then, continue with AI model templates in Studio, or use Brain Builder for AITRIOS for efficient AI model training.

Try your first object detection AI model:

Train an object detection AI model from Custom Vision ❯

Try an object detection model. Upload image data, label objects, and train the model.

Learn more about AI models:

Train AI models with Brain Builder for AITRIOS

If you want to evaluate your AI models with no-code tools, try Brain Builder for AITRIOS.

Browse AI model templates in Studio

Browse and download precompiled AI models, already trained and ready to deploy.
Note: Studio is not currently compatible with Console REST API V2 / Console V2. Support is planned in a future update.

Studio User Manual (for ConsoleV1) ❯

Deploy AI models

Deploy & run your AI models

Once your AI model is trained, convert it and deploy it to your edge device via Console. Then you can start to collect data and review the results.

In Console you can:

Convert your AI model for deployment
Deploy the model using your chosen configuration
Select a device to deploy your model on
Change inference settings
Run inference and collect data from your deployed AI model

Related docs

Deploy AI Model and Check Inference ❯

Deployment Guide for Custom Vision AI Model ❯

Deployment Guide for Keras AI Model ❯

Deployment Guide for PyTorch AI Model ❯

Try a sample application

Explore sample applications on GitHub

Explore sample applications to quickly test AI model deployment and inference on your edge device. Start with an example to see how the model runs directly on the device and responds to camera input in real time.

Zone detection

This sample application detects objects within configurable zones, combining a backend for Console interaction and a UI to display images with the detected objects and zone overlaps.

View the code on GitHub ❯

Basic visualization

This application implements object detection and classification, featuring a backend for Console interaction and a UI that displays images alongside detected objects.

View the code on GitHub ❯

Human detection

This solution implements human detection features including people counting and heatmaps. It includes a backend for Console interaction and a UI that displays images with detected objects using metadata from Console.

View the code on GitHub ❯

See more sample applications ❯

Next: Create your own solution

Build your custom edge application

Continue and create your own custom edge application using the Edge App SDK. Go one step further and create analytics and dashboards by fetching data from the Console REST API.

Related docs

Edge Application SDK ❯

Step-by-Step Guide to the Edge Application SDK ❯

Console REST API V2 Specification ❯

Console REST API V2 User Guide ❯