Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/marxmit7/memojify
deep learning method to turn facial expression into animated emoji
https://github.com/marxmit7/memojify
cnn deep-learning dlib emoji face keras python3 tensorflow
Last synced: about 4 hours ago
JSON representation
deep learning method to turn facial expression into animated emoji
- Host: GitHub
- URL: https://github.com/marxmit7/memojify
- Owner: marxmit7
- License: mit
- Created: 2018-05-29T16:19:46.000Z (over 6 years ago)
- Default Branch: master
- Last Pushed: 2021-01-17T14:13:14.000Z (almost 4 years ago)
- Last Synced: 2024-08-16T13:02:42.867Z (3 months ago)
- Topics: cnn, deep-learning, dlib, emoji, face, keras, python3, tensorflow
- Language: Python
- Homepage:
- Size: 807 KB
- Stars: 31
- Watchers: 1
- Forks: 5
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Memojify
emojify me nowInstall requirements
``` python -r requirements.txt ```
[Colab Link](https://colab.research.google.com/drive/19aggyFm_dCngSMb74Z-9zWFSTHCL6PEf#scrollTo=QQKe1zNw_fty)
## Files and their usage:
* ``` model.py```
* To train the model
* outputs the model & weight in model directory
* also outputs the graph having the logs of training
* arguments -e/--epochs
* pass the integer value of epochs* ```get_images_from_webcam.py```
* to create dataset using webcam (including all the required subfolders)
* It includes face alignment and face detection
* argparse -l/--label
* pass the argument like smile/wink/laugh according to your need* ``` utils.py ```
* contains useful required functions* ``` memojifier.py```
* It takes the stream of photos and converts the current emotion of the face into animated emojiCurrent Network Architecture
Result: with validation accuracy of 80%
### synopsis:
Currently model is unstable and overfits easily (:funny)
Try to train the model with as many as possible images and different environment
this project is currently in experimental stage and will be worked upon
## Contributions:-
Got some issue? raise an issue and if you have something to contribute then push a PR.