Emojify is a project built for purpose of learning computer vision and opencv tools. Basically it gives masks a filter based on the emotion of the user. The tech stack involves following
- Pytorch (for building model)
- dlib (for face detection)
- Open CV (almost entire project)
- And basic packages like Pandas, Numpy, Matplotlib, and PIL
emojify.mp4
-
Clone the project
git clone https://github.com/Navaneeth-Sharma/Emojify/
-
Move to Emojify directory
-
Download the requirements
pip install -r requirements.txt
-
For training
-
Generate Data using following commands
cd src/data ./generate_data.sh cv ..
-
Train the model
python train.py
-
-
For Running the system
-
Download the pretrained models (edit the model.pth for not downlading the emotion recognition model)
cd src/models ./download_models.sh cd ..
-
Run the main python file
python main.py
-