Avafly/YOLOv5-ncnn-OpenVINO-MNN-ONNXRuntime-OpenCV-CPP

YOLOv5 C++ inference implemented using multiple frameworks: ncnn, OpenVINO, MNN, ONNXRuntime, and OpenCV.

40
/ 100
Emerging

This project helps embedded systems engineers and robotics developers compare different inference engines for object detection. It takes pre-trained YOLOv5 models and processes images or live camera feeds, outputting detected objects with varying speed and resource usage depending on the chosen framework. It's designed for those who need to integrate real-time object detection into resource-constrained devices.

Use this if you are developing computer vision applications for edge devices and need to benchmark different object detection frameworks (like ncnn, OpenVINO, MNN) to find the most efficient one for your specific hardware and performance requirements.

Not ideal if you are an application developer looking for a high-level API to simply integrate object detection without needing to delve into C++ inference engine specifics or performance tuning.

edge-ai robotics embedded-systems computer-vision performance-optimization
No Package No Dependents
Maintenance 10 / 25
Adoption 6 / 25
Maturity 16 / 25
Community 8 / 25

How are scores calculated?

Stars

19

Forks

2

Language

C++

License

MIT

Last pushed

Jan 28, 2026

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/computer-vision/Avafly/YOLOv5-ncnn-OpenVINO-MNN-ONNXRuntime-OpenCV-CPP"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.