Overview

About Advantech Container Catalog

The Advantech Container Catalog delivers hardware-accelerated AI containers pre-integrated for seamless edge deployment. These containers abstract complexities like SDK setup, runtime compatibility, and toolchain dependencies—offering rapid development pathways for platforms such as the Qualcomm® QCS6490 SoC.

Container Overview

Pose Estimation on Qualcomm® Hexagon™ is a comprehensive container solution for running real-time pose estimation models on the QCS6490 platform. With full DSP acceleration, this container supports inference with models such as YOLOv8-Pose and HRNet, optimized for low-latency applications in the edge environment.

This container offers:

Dual Pose Estimation Workflows:
- Ultralytics Export: Use YOLOv8-native tools to export to TFLite for fast iteration and prototyping
- AI Hub Conversion: Import optimized HRNet models from Qualcomm’s Hugging Face repository for deployment-ready performance
Integrated Runtime Stack:
- Pre-installed support for QNN, SNPE, and LiteRT
- Includes GStreamer, OpenCV, and Python 3.10 for full pipeline development
Hardware-Accelerated Inference:
- INT8 inference on Hexagon™ DSP 770
- FP32 fallback and GPU acceleration via Adreno™ 643 GPU
Multi-Model Format Compatibility:
- Native support for .tflite, .dlc, and .so formats across supported runtimes
Preconfigured Scripts & Utilities:
- advantech-coe-model-export.sh and advantech-aihub-model-export.sh for exporting pose estimation models
- wise-bench.sh for validating AI environment and runtime compatibility
Ready for Edge AI Use Cases:
- Tailored for robotics, fitness tracking, motion capture, smart surveillance, AR/VR, and more
- Optimized for deployment on Advantech AOM-2721 with QCS6490 SoC
Seamless ROS Support:
- Plug-and-play compatibility with Qualcomm Robotics Reference Distro (ROS 1.3-ver.1.1) for robotic applications

Container Demo

Host Device Prerequisites

Component	Specification
Target Hardware	Advantech AOM-2721
SoC	Qualcomm® QCS6490
GPU	Adreno™ 643
DSP	Hexagon™ 770
Memory	8GB LPDDR5
Host OS	QCOM Robotics Reference Distro with ROS 1.3-ver.1.1

Container Environment Overview

Software Components on Container Image

Component	Version	Description
LiteRT	1.3.0	Provides QNN TFLite Delegate support for GPU and DSP acceleration
SNPE	2.29.0	Qualcomm’s Snapdragon Neural Processing Engine; optimized runtime for Snapdragon DSP/HTP
QNN	2.29.0	Qualcomm® Neural Network (QNN) runtime for executing quantized neural networks
GStreamer	1.20.7	Multimedia framework for building flexible audio/video pipelines
Python	3.10.12	Python runtime for building applications
OpenCV	4.11.0	Computer vision library for image and video processing

Quick Start Guide

For container quick start, including the docker-compose file and more, please refer to README.

Supported AI Capabilities

Vision Models

Model	Format	Note
YOLOv8 Detection	TFLite INT8	Downloaded from Ultralytics` official source and exported to TFLite using Ultralytics Python packages
YOLOv8 Segmentation	TFLite INT8	Downloaded from Ultralytics` official source and exported to TFLite using Ultralytics Python packages
YOLOv8 Pose Estimation	TFLite INT8	Downloaded from Ultralytics` official source and exported to TFLite using Ultralytics Python packages
Lightweight Face Detector	TFLite INT8	Converted using Qualcomm® AI Hub
FaceMap 3D Morphable Model	TFLite INT8	Converted using Qualcomm® AI Hub
DeepLabV3+ (MobileNet)	TFLite INT8	Converted using Qualcomm® AI Hub
DeepLabV3 (ResNet50)	SNPE DLC TFLite	Converted using Qualcomm® AI Hub
HRNet Pose Estimation (INT8)	TFLite INT8	Converted using Qualcomm® AI Hub
PoseNet (MobileNet V1)	TFLite	Converted using Qualcomm® AI Hub
MiDaS Depth Estimation	TFLite INT8	Converted using Qualcomm® AI Hub
MobileNet V2 (Quantized)	TFLite INT8	Converted using Qualcomm® AI Hub
Inception V3 (SNPE DLC)	SNPE DLC TFLite	Converted using Qualcomm® AI Hub
YAMNet (Audio Classification)	TFLite	Converted using Qualcomm® AI Hub
YOLO (Quantized)	TFLite INT8	Converted using Qualcomm® AI Hub

Language Models Recommendation

Model	Format	Note
Phi2	.so	Converted using Qualcomm's LLM Notebook for Phi-2
Tinyllama	.so	Converted using Qualcomm's LLM Notebook for Tinyllama
Meta Llama 3.2 1B	.so	Converted using Qualcomm's LLM Notebook for Meta Llama 3.2 1B

Supported AI Model Formats

Runtime	Format	Compatible Versions
QNN	.so	2.29.0
SNPE	.dlc	2.29.0
LiteRT	.tflite	1.3.0

Hardware Acceleration Support

Accelerator	Support Level	Compatible Libraries
GPU	FP32	QNN, SNPE, LiteRT
DSP	INT8	QNN, SNPE, LiteRT

Best Practices

Prefer INT8 quantized models for DSP acceleration
Ensure fixed batch sizes when converting models
Use lower GST_DEBUG levels for stable multimedia handling
Always validate exported models on-device after deployment