Skip to content

Latest commit

 

History

History

YOLOv3-Training-Snowman-Detector

README.md

Training YOLOv3 : Deep Learning based Custom Object Detector

This repository contains the code for Training YOLOv3 : Deep Learning based Custom Object Detector blog post.

download

  1. Install awscli

sudo pip3 install awscli

  1. Get the relevant OpenImages files needed to locate images of our interest

wget https://storage.googleapis.com/openimages/2018_04/class-descriptions-boxable.csv

wget https://storage.googleapis.com/openimages/2018_04/train/train-annotations-bbox.csv

  1. Download the images from OpenImagesV4

python3 getDataFromOpenImages_snowman.py

  1. Create the train-test split

python3 splitTrainAndTest.py /data-ssd/sunita/snowman/JPEGImages

Give the correct path to the data JPEGImages folder. The 'labels' folder should be in the same directory as the JPEGImages folder.

  1. Install Darknet and compile it.
cd ~
git clone https://github.com/pjreddie/darknet
cd darknet
make
  1. Get the pretrained model

wget https://pjreddie.com/media/files/darknet53.conv.74 -O ~/darknet/darknet53.conv.74

  1. Fill in correct paths in the darknet.data file

  2. Start the training as below, by giving the correct paths to all the files being used as arguments

cd ~/darknet

./darknet detector train /path/to/snowman/darknet.data /path/to/snowman/darknet-yolov3.cfg ./darknet53.conv.74 > /path/to/snowman/train.log

  1. Give the correct path to the modelConfiguration and modelWeights files in object_detection_yolo.py and test any image or video for snowman detection, e.g.

python3 object_detection_yolo.py --image=snowmanImage.jpg

AI Courses by OpenCV

Want to become an expert in AI? AI Courses by OpenCV is a great place to start.