README.md

Image Classification of Human Facial Expression using Custom CNN Architecture

Objective :

To build image classifier to classify human facial expression into happy and sad faces.

Data Set :

1.Images for this data is Scrap from google using `All Images Download` Chrome Extension
2.Image are Scrap for Happy Faces and Sad Faces

Deep Learning Model:

1. Train 70% data, 20% Validation data, 10% Test Data
2. Architecture : 3 CNN Layers with Max pooling and 2 Dense Layer
3. Optimizer = Adam, Losses= BinaryCrossEntropy, Epochs=20 , Batch Size=20
4. Data pipeline using tf.keras.utils
5. Tensorflow version ==2.9.1