Emotion_Prediction_Kaggle_CNN

Facial Expression Prediction with CNN Model Architecture making use of Transfer Learning, Data Augmentation, Regularization Techniques.

i> The Kaggle Notebook file makes use of the CNN model architecture combined with Transfer Learning(Resnet-9).

ii> The model present in the notebook is a prediction model which tries to label a multi class dataset accurately.

iii> The model makes use of Data Augmentation Techniques and Regularization techniques to generate a accuracy of 83%.

iv> The model also makes use of Residual Blocks and Batch Normalisation Techniques.

The dataset used here is Facial Expression Recog Image Ver of (FERC)Dataset . ⬇️

The dataset contains 7 labels namely:

i> ['anger'] 😠

ii> ['fear'] 😨

iii> ['surprise'] 😍

iv> ['happiness'] 😃

v> ['sadness'] 😢

vi> ['neutral'] 😐

vii> ['disgust'] 🤮

Please contribute using GitHub Flow . Create a branch, add commits, and open a pull request.