- Install Anaconda(
Python 3.7
) & Tensorflow GPU (I used 1.15 TF
) - First Download Object Detecion official code from here
- Then Download this Repository via GIT or zip.
- Download Training.zip
- In my Repository extract `Training.zip.
- Now copy all the files & folders from my repository.
- Go to the
master/research/object_detection
and paste all files & folders of my repository to there. - Open
face_mask_detection.ipynb
in Jupyter and give the image in which you want to detect Mask(Mentioned in code).
- This is already prepared which I given to you. But behind of these I created my own image dataset, then did Image Labelling,Model configuration.
- After that I used
Faster_rcnn_v2
model and I trained it around 4 hours. At the end i got a loss of0.053
. - Model is trained os NVIDA 920MX 2GB(used TensorFlow-GPU with CUDA)
- So guys I given lots of efforts to complete this project. So please Star🌟 my repository and follow me.
- If you have any queries the you can ask me on my Insta Community Page