A deep learning approach to create passport size images by applying different deep learning method.
- Remove background from image
- fixed image orientation
- face alignment
- croping image from neck to top in specific ration footage**
Before fixing orientation | After making in portrait |
Before fixing alignment | After fixing alignment |
Before removing the background | After removing the background |
Before making passport size | After making passport size |
- Python == 3.8
python -m venv venv
.\venv\Scripts\activate
- If any problem for scripts activation
- Execute following command in administration mode
Set-ExecutionPolicy Unrestricted -Force
- Later you can revert the change
Set-ExecutionPolicy restricted -Force
- Execute following command in administration mode
python -m venv venv
source venv/bin/activate
- Library Install
tensorflow version 1.x
onnx
onnxruntime
mtcnn
- download weight from referece project link.
It will generate the output file in temp/ folder
python inference.py
-
A Trimap-Free Solution for Portrait Matting in Real Time under Changing Scenes
-
Smile detection using keras Smiler
-
Pose estimation and matching with TensorFlow lite PoseNet model. Link