model.py

import tensorflow as tf
import pandas as pd
from tensorflow import keras
from keras.models import Sequential, Model
from keras.layers import Dense, Conv2D , MaxPool2D , Flatten , Dropout , Concatenate, GlobalAveragePooling2D, Multiply
from keras.applications.vgg19 import VGG19
from keras.applications.resnet_v2 import ResNet101V2
from keras.applications import ResNet50


def build_VGG19(IMG_SIZE):
    pre_trained_model = VGG19(input_shape= (IMG_SIZE, IMG_SIZE, 3), include_top=False, weights="imagenet")

    for layer in pre_trained_model.layers[:19]:
        layer.trainable = False
        

    model = Sequential([
    pre_trained_model,
    MaxPool2D((2,2) , strides = 2),
    Flatten(),
    Dense(5 , activation='softmax')])
    
    return model

def build_ResNet101V2(IMG_SIZE):
    pre_train_model = ResNet101V2(input_shape = (IMG_SIZE, IMG_SIZE, 3), include_top=False, weights="imagenet" )
    
    model = Sequential([
    pre_train_model,
    MaxPool2D((2,2) , strides = 2),
    Flatten(),
    Dense(5 , activation='softmax')])
    
    return model

def build_RetinaNet(IMG_SIZE):
    input_tensor = keras.Input(shape=(IMG_SIZE,IMG_SIZE,3))
    feature_extractor = ResNet50(include_top=False, weights='imagenet', input_shaoe=input_tensor)


# Define the classification sub-network
    classification = Conv2D(filters=9, kernel_size=(3,3), activation='relu')(feature_extractor.output)
    #classification = Flatten()(classification)
    #classification = Dense(units=9, activation='sigmoid')(classification) # Remove the regression subnetwork


# Define the regression sub-network
    regression = Conv2D(filters=36, kernel_size=(3,3), activation='relu')(feature_extractor.output)
    #regression = Flatten()(regression) 
    #regression = Dense(units=4)(regression)  # Remove the classification subnetwork
    
    # Concatenate the outputs from the classification and regression sub-networks
    concat = Concatenate()([classification, regression])
   
    # retinanet for classification
     
    pooling = MaxPool2D((2,2) , strides = 2)(concat)
    flatten = Flatten()(pooling)
    output = Dense(5 , activation='softmax')(flatten) 
    
    # Define the final model
    model = Model(inputs=input_tensor, outputs=output)
    
    return model

def build_ResNet101V2_Attention(IMG_SIZE):
    input_tensor = keras.Input(shape=(IMG_SIZE, IMG_SIZE, 3))
    feature_extractor = ResNet101V2(include_top=False, weights='imagenet', input_tensor=input_tensor)
  # Define a spatial attention layer
    attention = Conv2D(1, (1, 1), activation='sigmoid')(feature_extractor.output)

# Multiply the original features with attention weights
    attended_features = Multiply()([feature_extractor.output, attention])
    
    #attended_features = AttentionLayer.AttentionLayer()(feature_extractor.output)
    pooling = MaxPool2D((2,2) , strides = 2)(attended_features)
    flat = Flatten()(pooling)
    dense = Dense(5, activation='softmax')(flat)
  
    model = Model (inputs= input_tensor, outputs=dense)
    return model