Add files via upload

Author: back-kh

models/cvt_model.py

@@ -0,0 +1,16 @@
+import tensorflow as tf
+from tensorflow.keras import layers
+from tensorflow.keras.applications import VGG16  # Placeholder for ViT
+
+def build_vit_model(num_classes):
+    inputs = tf.keras.Input(shape=(224, 224, 3))
+    base_model = VGG16(include_top=False, input_tensor=inputs, weights='imagenet')  # Use a ViT implementation here
+    base_model.trainable = True
+
+    x = layers.Flatten()(base_model.output)
+    x = layers.Dropout(0.5)(x)
+    outputs = layers.Dense(num_classes, activation='softmax')(x)
+
+    model = tf.keras.Model(inputs, outputs)
+    model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])
+    return model

models/efficientnet_model.py

@@ -0,0 +1,17 @@
+import tensorflow as tf
+from tensorflow.keras import layers
+from tensorflow.keras.applications import EfficientNetB0
+
+def build_efficientnet_model(num_classes):
+    inputs = tf.keras.Input(shape=(224, 224, 3))
+    base_model = EfficientNetB0(include_top=False, input_tensor=inputs, weights='imagenet')
+    base_model.trainable = True
+
+    x = layers.GlobalAveragePooling2D()(base_model.output)
+    x = layers.BatchNormalization()(x)
+    x = layers.Dropout(0.4)(x)
+    outputs = layers.Dense(num_classes, activation="softmax")(x)
+
+    model = tf.keras.Model(inputs, outputs)
+    model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])
+    return model

models/model.py

@@ -0,0 +1,55 @@
+import tensorflow as tf
+from tensorflow.keras import layers
+from tensorflow.keras.applications import EfficientNetB0, ResNet50
+from tensorflow.keras.applications import ResNet50, VGG16
+from tensorflow.keras.layers import Input, Dense, GlobalAveragePooling2D
+
+# EfficientNet model
+def build_efficientnet_model(num_classes):
+    inputs = Input(shape=(224, 224, 3))
+    base_model = EfficientNetB0(include_top=False, input_tensor=inputs, weights='imagenet')
+    base_model.trainable = True
+
+    x = GlobalAveragePooling2D(name="avg_pool")(base_model.output)
+    x = layers.BatchNormalization()(x)
+    x = layers.Dropout(0.4, name="top_dropout")(x)
+    outputs = Dense(num_classes, activation="softmax", name="pred")(x)
+
+    model = tf.keras.Model(inputs, outputs)
+    model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])
+    return model
+
+# ResNet model
+def build_resnet_model(num_classes):
+    inputs = Input(shape=(224, 224, 3))
+    base_model = ResNet50(include_top=False, input_tensor=inputs, weights='imagenet')
+    base_model.trainable = True
+
+    x = GlobalAveragePooling2D(name="avg_pool")(base_model.output)
+    x = layers.BatchNormalization()(x)
+    x = layers.Dropout(0.4, name="top_dropout")(x)
+    outputs = Dense(num_classes, activation="softmax", name="pred")(x)
+
+    model = tf.keras.Model(inputs, outputs)
+    model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])
+    return model
+
+# Vision Transformer model (ViT)
+def build_vit_model(num_classes):
+    from tensorflow.keras.layers import Dense, Flatten, Dropout
+    from tensorflow.keras.applications import VGG16
+
+    # Here, we can replace VGG16 with an actual ViT implementation or a library that supports it
+    # For demonstration, I will use a placeholder approach
+    # Note: You need to implement or import a ViT architecture from a compatible library
+    inputs = Input(shape=(224, 224, 3))
+    base_model = VGG16(include_top=False, input_tensor=inputs, weights='imagenet')
+    base_model.trainable = True
+
+    x = Flatten()(base_model.output)
+    x = Dropout(0.5)(x)
+    outputs = Dense(num_classes, activation='softmax')(x)
+
+    model = tf.keras.Model(inputs, outputs)
+    model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])
+    return model

models/resnet_model.py

@@ -0,0 +1,17 @@
+import tensorflow as tf
+from tensorflow.keras import layers
+from tensorflow.keras.applications import ResNet50
+
+def build_resnet_model(num_classes):
+    inputs = tf.keras.Input(shape=(224, 224, 3))
+    base_model = ResNet50(include_top=False, input_tensor=inputs, weights='imagenet')
+    base_model.trainable = True
+
+    x = layers.GlobalAveragePooling2D()(base_model.output)
+    x = layers.BatchNormalization()(x)
+    x = layers.Dropout(0.4)(x)
+    outputs = layers.Dense(num_classes, activation="softmax")(x)
+
+    model = tf.keras.Model(inputs, outputs)
+    model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])
+    return model

models/swin_model.py

@@ -0,0 +1,16 @@
+import tensorflow as tf
+from tensorflow.keras import layers
+
+# Placeholder for Swin Transformer implementation
+def build_swin_model(num_classes):
+    inputs = tf.keras.Input(shape=(224, 224, 3))
+    # Implement Swin Transformer architecture here
+
+    # Placeholder output for demonstration
+    x = layers.Flatten()(inputs)
+    x = layers.Dropout(0.5)(x)
+    outputs = layers.Dense(num_classes, activation='softmax')(x)
+
+    model = tf.keras.Model(inputs, outputs)
+    model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])
+    return model