Real_Time_Emotion_Recognition

Overview:

The task here is to detect facial expression from real time video. Mainly 7 classes of emotion can be detected like angry, disgust, scared, happy, sad, surprised and neutral.

Requirements:

cv2

keras

tensorflow

imutils

Usage:

1. To run the pretrained model use:

python RealTimeEmotionDetector.py

Accuracy: 67%

2. To train the model from scratch use:

python TrainModelClassifier.py

Dataset:

Used this dataset

Download it and place in the same folder where the project files are kept.