Skip to content

Lab-of-Infinity/Facial-Expression-Image-Classification-Project-using-Deep-CNN

Repository files navigation

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.

image

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 

Releases

No releases published

Packages

No packages published