Overview
About Advantech Container Catalog (ACC)
Advantech Container Catalog is a comprehensive collection of ready-to-use, containerized software packages designed to accelerate the development and deployment of Edge AI applications. By offering pre-integrated solutions optimized for embedded hardware, ACC simplifies the challenge of software-hardware compatibility, especially in GPU/NPU-accelerated environments.
| Feature / Benefit | Description |
|---|---|
| Accelerated Edge AI Development | Ready-to-use containerized solutions for faster prototyping and deployment |
| Hardware Compatible | Eliminates embedded hardware and software package incompatibility |
| GPU/NPU Access Ready | Supports passthrough for efficient hardware acceleration |
| Model Conversion & Optimization | Built-in AI model quantization and format conversion support |
| Optimized for CV & LLM Applications | Pre-optimized containers for computer vision and large language models |
Container Overview
MemryX NPU PPE Detection showcases a Personal Protective Equipment (PPE) detection application using the YOLOv8n model with 512x512 input resolution. The system is designed to predict and classify individuals based on their safety compliance in real-time video streams or images. The model identifies three categories:
- Safety: The person is wearing both a helmet and a vest.
- Half-Safety: The person is wearing either a helmet or a vest, but not both.
- Non-Safety: The person is wearing neither a helmet nor a vest.
By leveraging advanced object detection capabilities, the application helps monitor and enforce safety protocols in industrial or construction environments, ensuring that personnel adhere to required safety standards.
Container Demo
Key Features of the MemryX MX3 Accelerator
- High-Efficiency AI Inference : Utilizes a unique dataflow architecture with at-memory computing, enabling fast and power-efficient AI model inference.
- Seamless Integration : Connects to host systems via standard PCIe or USB interfaces, supporting x86, ARM, and RISC-V processors on both Windows and Linux platforms.
- Full AI Offloading : Handles all AI inference on-chip, requiring the host only for pre- and post-processing, which simplifies software development and deployment.
Requirements
- Ensure your host machine is equipped with a MemryX MX3 module.
- Follow the installation instructions at MemryX Developer Hub to install the MX3 driver. Verify that the driver is successfully installed on your device before proceeding.
- X11 Display Support: Facilitates seamless rendering of model outputs directly to a host display from inside the container environment.
- (Optional) USB Camera Access:
The application can utilize a connected USB camera for live streaming PPE detection. Ensure your camera is plugged in and accessible to the container for real-time video inference.
Host Device Prerequisites
| Item | Specification |
|---|---|
| Compatible Hardware | Advantech devices accelerated by MemryX MX Module - refer to Compatible hardware |
| MemryX MX Module | MemryX MX module - refer to MemryX M.2 Module |
| Host OS | Ubuntu 24.04 |
| Required Software packages | *refer to below |
Required Software Packages on Host Device
| Component | Version | Description |
|---|---|---|
| MemryX MX Driver | 1.2.0 | MemryX MX module driver |
Container Environment Overview
Software Components on Container Image
| Component | Version | Description |
|---|---|---|
| MemryX SDK | 1.2.0 | MemryX development and inference SDK |
| CMake | 3.28.3 | Build system tool |
| OpenCV | 4.6.0 | Computer vision library |
| Qt | 5.15.13 | Cross-platform GUI framework |
| ONNX Runtime | 1.18.1 | Cross-platform inference engine |
Container Quick Start Guide
For container quick start, including docker-compose file, and more, please refer to Advantech EdgeSync Container Repository
Key Features
-
Edge-Optimized AI Inference
Efficiently runs AI models directly on edge devices, minimizing cloud dependency. -
Framework & Model Agnostic
Supports TensorFlow, PyTorch, ONNX, and more – plug-and-play with most AI models. -
High Efficiency
Industry-leading TOPS/Watt performance for fanless and power-constrained systems. -
Compact Form Factors
Available in M.2 and PCIe for easy integration into IPCs, gateways, and embedded systems.
Example Use Cases
| Application | Description |
|---|---|
| PPE Detection | Real-time safety gear compliance monitoring |
| Anomaly Detection | Machine fault or behavioral outlier detection |
| Smart Cameras | On-device object/person detection & tracking |
| Robotics | Vision-based navigation and decision-making |
| Retail Analytics | Customer behavior analysis at the edge |
About MemryX
MemryX is a U.S.-based AI semiconductor company developing edge-native Neural Processing Units (NPUs) that deliver high-performance, low-latency, and energy-efficient AI inferencing.
Partnership with Advantech
MemryX NPUs are integrated with Advantech's industrial PCs, enabling:
- Turnkey Edge AI solutions
- Low-latency real-time inference
- Cost-effective AIoT deployment at scale