Humans are used to non verbal communication. The emotions expressed increases the clarity of any thoughts and ideas. It becoms quite interesting when a computer can capture this complex feature of humans, ie emotions. This topic talks about building a model which can detect an emotion from an image. There key points to be followed are:
-
Data gathering and augmentation
The dataset taken was "fer2013". It can be downloaded through the link "https://github.com/npinto/fer2013". Image augmentation was performed on this data.
-
Model building
The model architecture consists of CNN Layer, Max Pooling, Flatten and Dropout Layers.
-
Training
The model was trained by using variants of above layers mentioned in model building and by varying hyperparameters. The best model was able to achieve 60.1% of validation accuracy.
-
Testing
The model was tested with sample images. It can be seen below:
Refer to the notebook /Emotion_Detection.ipynb.
I have trained an emotion detection model and put its trained weights at /Models
To train your own emotion detection model, Refer to the notebook /facial_emotion_recognition.ipynb
Run pip install -r requirements.txt
python Emotion_Detection.py