← Back

ofxYolov2

github.com/TetsuakiBaba/ofxYolov2 · @TetsuakiBaba
OF 0.10 Machine Learning

Yolov2 for Openframworks addon

16 4 Last commit 7y ago Created 8y ago

Forks

Forks pushed to more recently than TetsuakiBaba/ofxYolov2 itself, plus any admin-confirmed fork with commits of its own — worth checking if you're hitting issues with the original.

technologiestiftung/ofxYolov2 by @technologiestiftung · ★ 0 · pushed 7y ago

README

ofxYolov2

This project was merged into ofxOpenCvDnnObjectDetection ( https://github.com/TetsuakiBaba/ofxOpenCvDnnObjectDetection ), therefore, it is not maintained any more.



Description

ofxYolov2 is a YOLO(You Look Only Once) addon for openframeworks. Please refer YOLO official website for more details of YOLO. https://pjreddie.com/darknet/yolo/ If you have a Nvidia GPU(s) on your PC, I just recommend to use ofxDarknet instead of this addon, https://github.com/mrzl/ofxDarknet.

OpenCV v.3.3.1 or upper includes Yolov2 library ( https://docs.opencv.org/3.4.1/da/d9d/tutorial_dnn_yolo.html ) in their own package. Therefore I designed ofxYolov2 with OpenCV v.3.3.1 or upper version.

I created this addon for beginers or learner who are going to try object detection with Yolo on OF. Attention: it runs on OpenCL.

Usage

See Examples

single_image

single_image screenshot Detection example for single image.

camera

Realtime Detection example for webcam video.

annotation

annotation screenshot Simple annotation example to export a yolo format.

Install

1. Install OpenCV by brew.



$ brew install opencv

2. confirm your opencv version

Based on OpenCV version, Please re-write ofxYolov2/addon_config.mk file. End of the file, there are some description about opencv path for your system. Please re-write the version according to your OpenCV version. Current (2018/6/15) is 3.4.1_5, If you find other version on your directory tree, /usr/local/Cellar/opencv , please do not forget fill in the correct version on the addon_config.mk file. You can get your current opencv version by typing below command.

$ ls /usr/local/Cellar/opencv

3. Download cfg and weights file

Download cfg and weights file to each examples directories. You can find download shell script on the top of ofxYolov2 directory.

$ cd ofxYolov2 $ sh getYolov2Weights.sh

4. Edit yolov2-tiny.cfg

Edit "thresh .6" to "thresh .01" on each yolov2-tiny.cfg.

5. Update each examples by project generator, then run.

Example/camera, single_image or annotation by project generator and add only ofxYolov2 as an addon. Then update it. _Attention Please do not include OpenCV addon in OF. Be sure to add only ofxYolov2 addon by projectgenerator._

Compatibility

Licence

MIT

Author

TetsuakiBaba

Reference

1. How to implement OpenCV Yolo on Openframeworks: https://qiita.com/buchayaty/items/4020100f531c07418f38 2. YOLO: https://pjreddie.com/darknet/yolo/

Get this addon

HTTPS
SSH