- Python 3, OpenCV 3 or 4, Tensorflow 1 or 2
- To install the required packages, run
pip install -r requirements.txt
.
The repository is currently compatible with tensorflow-2.0
and makes use of the Keras API using the tensorflow.keras
library.
-
If you want to train this model or train after making changes to the model, use
python emotions.py --mode train
. -
The folder structure is of the form:
Tensorflow:- data (folder)
emotions.py
(file)haarcascade_frontalface_default.xml
(file)model.h5
(file)
-
This implementation by default detects emotions on all faces in the webcam feed.
-
With a simple 4-layer CNN, the test accuracy peaked at around 50 epochs at an accuracy of 63.2%.
-
First, we use haar cascade to detect faces in each frame of the webcam feed.
-
The region of image containing the face is resized to 48x48 and is passed as input to the ConvNet.
-
The network outputs a list of softmax scores for the seven classes.
-
The emotion with maximum score is displayed on the screen.