The goal of this project is to use a CNN model to analyze live video input of multiple facial expressions and create a generealization of mood among people. Once a consensus is made a playlist with metadata matching the expression/mood will play.
Newer implementation with a deeper network, data augmentation and learning rate schedule. Current itteration:
#First layer to normalize the RGB chanels model.add(layers.Rescaling(1./255))
model.add(data_augmentation)
#4 layer stack for convlutional block
model.add(layers.Conv2D(64, 3, padding='same',activation='relu')) model.add(layers.MaxPooling2D())
model.add(layers.Conv2D(128, 3, padding='same',activation='relu')) model.add(layers.BatchNormalization()) model.add(layers.Conv2D(128, 3, padding='same',activation='relu')) model.add(layers.MaxPooling2D()) model.add(layers.Dropout(0.20))
model.add(layers.Conv2D(256, 3, padding='same',activation='relu')) model.add(layers.BatchNormalization()) model.add(layers.MaxPooling2D()) model.add(layers.Dropout(0.20))
model.add(layers.Conv2D(512, 3, padding='same',activation='relu')) model.add(layers.BatchNormalization()) model.add(layers.MaxPooling2D()) model.add(layers.Dropout(0.20))
#Flatten for dense layer model.add(layers.Flatten()) model.add(layers.BatchNormalization()) model.add(layers.Dropout(.20, input_shape=(3,))) model.add(layers.Dense(128, activation='relu')) model.add(layers.Dropout(.20)) model.add(layers.Dense(num_classes,activation = 'softmax'))
