Skip to content

This repository aims to create a mobile image segmentation model and incorporate background according to the human emotion

License

Notifications You must be signed in to change notification settings

karan469/Smart-Filter

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Smart Filter

This is the repository for Smart-Filter, a deep learning open source web app. This repository provides end-to-end pipeline from model architecture of face detection using transfer learning to deployment of a web application using Docker container.

Visit the official repository webpage: https://karan469.github.io/Smart-Filter/

Web app soon to be deployed.

Requirements

Caution: Specified softwares must be installed in their correct versions. Mentioned versions are compatible with each other.

Subdirectories

Deployment/
Web application for deploying a trained set of models on Flask using Docker containers. To read more about Docker, visit official Docker documentation: https://docs.docker.com/

  • templates/
  • uploads/
  • app.py
  • detectron.py
  • smile.py
  • utils.py
  • facedetector.py
  • requirements.txt

training/
Contains training modules for semantic segmentation, face detection and smile detection. iPython notebooks contains relevant model architectures.

  • Face Detection using Detectron2/
  • Final Ensemble/
  • Key Points Detection/
  • Semantic Segmentation/
  • Smile Detection

results/
Contains demo images as examples of working prototype of final web application consisting of features such as custom background and caption writing according to facial features such as smile.

Contact

For more information visit the official documentations of different frameworks.
For any queries, email me at mailto://tkaran.iitd@gmail.com

About

This repository aims to create a mobile image segmentation model and incorporate background according to the human emotion

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages