Home - up-division/up-ai GitHub Wiki

Welcome to the AAEON UP AI toolkit

Overview

The AAEON UP AI Toolkit is a comprehensive software solution designed to accelerate AI development and deployment on AAEON edge computing platforms. It provides seamless integration of AI frameworks and advanced hardware accelerators, enabling real-time object detection, natural language processing and other functions. Whether you’re working with vision-based applications or chatbots, the toolkit delivers high-performance capabilities tailored for a wide range of use cases, from industrial automation to smart environments. It supports a variety of AI accelerators, such as Hailo 8, Intel NPU, and Nvidia GPU, and is compatible with mainstream operating systems, providing a user-friendly interface for quick start and environment configuration.

System Requirements

Hardware Compatibility

UP Squared
UP Squared Pro
UP Squared V2
UP Xtreme i11
UP Xtreme i12
UP Xtreme i14
UP Squared 6000
UP 4000
UP 7000
UP 7100 Squared
UP Squared Pro 7000
UP Squared i12

Supported AI Accelerators

Hailo 8
Intel NPU
Intel GPU
Nvidia GPU

Operating System Support

AI Accelerator	Windows 10 21H2 LTSC	Windows 11 24H2 LTSC	Ubuntu 22.04	Ubuntu 24.04
Hailo 8	✅	✅	✅	❌
Intel NPU	❌	✅	✅	✅
Intel GPU	✅	✅	✅	✅
Nvidia GPU	✅	✅	✅	✅

Quick Start Guide

Download latest version v1.0.0 and unzip it to up-ai folder

Windows Installation

Navigate to installation directory: up-ai
Run prepare.bat
- System reboot required after installation
- Internet connection required
Launch application with Start_app.bat
- Follow prompts to select demo type and hardware

Linux Installation

Navigate to installation directory:
```
cd up-ai
```
Give prepare.sh and start_app.sh execution permissions
```
chmod +x prepare.sh start_app.sh
```
Run prepare.sh
- Select option "2" for automatic installation
- System reboot required after installation
- Internet connection required
Launch application with start_app.sh
- Follow prompts to select demo type and hardware
```
./start_app.sh
```

AI Examples

Chatbot

Our chatbot leverages advanced Natural Language Processing (NLP) technology to:

Process and understand user text input
Generate contextually relevant responses
Enable natural conversational flow
Handle complex semantic and grammatical structures

chatbot

Object Detection

Real-time object detection capabilities include:

Live video and camera feed processing
Multi-object detection and tracking
Support for various object classes (people, vehicles, animals, etc.)
Real-time performance optimization

Note: The default camera index is set to 0, which corresponds to the first camera device.

⭐ You can press 'C' to view CPU usage, press 'M' to view memory usage, or press 'A' to view both CPU and memory usage together.