Skip to content

Latest commit

 

History

History
69 lines (47 loc) · 1.79 KB

README.md

File metadata and controls

69 lines (47 loc) · 1.79 KB
Raw

OpenVINO-Deploy

OpenVINO-Deploy aims to implement popular AI models' inference deployment with OpenVINO written in C++, with a primary focus on object detection algorithms.

Test Environment

  1. OpenVINO 2023.3.0
  2. OpenCV 4.2.0
  3. Ubuntu 20.04

Models

Following models are implemented.

Name Description
yolov3 A fast and accurate real-time object detection algorithm that can detect multiple objects in an image or video stream with high efficiency.
yolov5 A fast, accurate, and user-friendly real-time object detection algorithm.
yolov7 A cutting-edge real-time object detection model delivering exceptional accuracy and speed.
yolov8 An enhanced YOLO object detection model with improved performance and flexibility.
yolov9 Paper yolov9 Implement, Learning What You Want to Learn Using Programmable Gradient Information.
rt-detr RT-DETR is a real-time object detection model that leverages transformers for efficient and accurate performance.

Quick Start

  1. Clone this repo
git clone https://github.com/wxxz975/OpenVINO-Deploy.git
  1. Create Build Directory
mkdir build && cd build
  1. Build All
cmake .. && make -j$(nproc)

Then, you will find executable files for demos in this build directory.

The onnx models exported by the experiment can be found on BaiduCloud

TODO List

  • yolov3
  • yolov3-spp
  • yolov3-tiny
  • yolop
  • yolov4
  • yolov5
  • yolov7
  • yolov8
  • yolov9
  • detr
  • rt-detr
  • ssd
  • faster-rcnn