Adding model code

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+xray_image_classifier_model.keras filter=lfs diff=lfs merge=lfs -text

README

@@ -0,0 +1 @@
+poetry run python train_model.py

augment_images.py

@@ -0,0 +1,64 @@
+import os
+from PIL import Image
+from tensorflow.keras.preprocessing.image import ImageDataGenerator
+
+# Set paths
+base_dir = 'data/chest_xray'
+val_dir = os.path.join(base_dir, 'val')
+normal_class_dir = os.path.join(val_dir, 'NORMAL')
+pneumonia_class_dir = os.path.join(val_dir, 'PNEUMONIA')
+
+
+def augment_images(class_directory, num_augmented_images):
+    datagen = ImageDataGenerator(
+        rescale=1. / 255,
+        rotation_range=20,
+        width_shift_range=0.2,
+        height_shift_range=0.2,
+        shear_range=0.2,
+        zoom_range=0.2,
+        horizontal_flip=True,
+        fill_mode='nearest'
+    )
+
+    generator = datagen.flow_from_directory(
+        directory=os.path.dirname(class_directory),  # Parent directory
+        target_size=(150, 150),
+        batch_size=1,
+        class_mode=None,
+        shuffle=False,
+        classes=[os.path.basename(class_directory)]  # Specify class if using subdirectory
+    )
+
+    print(f"Found {generator.samples} images in {class_directory}")
+
+    if generator.samples == 0:
+        print("No images found in the directory.")
+        return
+
+    count = 0
+
+    while count < num_augmented_images:
+        try:
+            img_batch = generator.__next__()  # Use __next__() to get image batch
+            img = (img_batch[0] * 255).astype('uint8')  # Extract the first image in the batch
+            img_pil = Image.fromarray(img)
+            img_path = os.path.join(class_directory, f"augmented_{count}.png")
+            img_pil.save(img_path)
+            count += 1
+        except StopIteration:
+            print("No more images to generate.")
+            break
+
+    print(f"Total augmented images created: {count}")
+
+
+# Number of augmented images to generate
+num_augmented_images_normal = 2944 - 3875  # This should be a negative number since NORMAL is already balanced
+num_augmented_images_pneumonia = 2944 - 1171  # To match the number of NORMAL images
+
+# Generate augmented images for the NORMAL class
+augment_images(normal_class_dir, max(num_augmented_images_normal, 0))
+
+# Generate augmented images for the PNEUMONIA class
+augment_images(pneumonia_class_dir, num_augmented_images_pneumonia)

data_check.py

@@ -0,0 +1,55 @@
+import os
+from PIL import Image
+
+# Define the data directories
+base_dir = 'data/chest_xray'
+train_dir = os.path.join(base_dir, 'train')
+val_dir = os.path.join(base_dir, 'val')
+
+
+# Function to count images in a specific category (e.g., NORMAL, PNEUMONIA)
+def count_images(directory, category):
+    category_dir = os.path.join(directory, category)
+    count = 0
+    for root, dirs, files in os.walk(category_dir):
+        count += len([f for f in files if f.endswith(('.jpg', '.jpeg', '.png'))])
+    return count
+
+
+# Function to check for corrupted images in a specific category
+def check_corrupted_images(directory, category):
+    category_dir = os.path.join(directory, category)
+    corrupted_files = []
+    for root, dirs, files in os.walk(category_dir):
+        for file in files:
+            if file.endswith(('.jpg', '.jpeg', '.png')):
+                try:
+                    img = Image.open(os.path.join(root, file))
+                    img.verify()  # Check if the image can be opened and is not corrupted
+                except (IOError, SyntaxError) as e:
+                    corrupted_files.append(os.path.join(root, file))
+    return corrupted_files
+
+
+# Count images in the train and validation sets
+train_normal_count = count_images(train_dir, 'NORMAL')
+train_pneumonia_count = count_images(train_dir, 'PNEUMONIA')
+val_normal_count = count_images(val_dir, 'NORMAL')
+val_pneumonia_count = count_images(val_dir, 'PNEUMONIA')
+
+# Check for corrupted images in the train and validation sets
+train_normal_corrupted = check_corrupted_images(train_dir, 'NORMAL')
+train_pneumonia_corrupted = check_corrupted_images(train_dir, 'PNEUMONIA')
+val_normal_corrupted = check_corrupted_images(val_dir, 'NORMAL')
+val_pneumonia_corrupted = check_corrupted_images(val_dir, 'PNEUMONIA')
+
+# Print the results
+print(f"Training NORMAL images: {train_normal_count}")
+print(f"Training PNEUMONIA images: {train_pneumonia_count}")
+print(f"Validation NORMAL images: {val_normal_count}")
+print(f"Validation PNEUMONIA images: {val_pneumonia_count}")
+
+print(f"Corrupted images in training NORMAL: {train_normal_corrupted}")
+print(f"Corrupted images in training PNEUMONIA: {train_pneumonia_corrupted}")
+print(f"Corrupted images in validation NORMAL: {val_normal_corrupted}")
+print(f"Corrupted images in validation PNEUMONIA: {val_pneumonia_corrupted}")

evaluate_model.py

@@ -0,0 +1,21 @@
+import os
+import tensorflow as tf
+from tensorflow.keras.preprocessing.image import ImageDataGenerator
+from tensorflow.keras.models import load_model
+
+base_dir = 'data/chest_xray'
+val_dir = os.path.join(base_dir, 'val')
+
+val_datagen = ImageDataGenerator(rescale=1./255)
+val_generator = val_datagen.flow_from_directory(
+    val_dir,
+    target_size=(150, 150),
+    batch_size=32,
+    class_mode='binary'
+)
+
+model = load_model('xray_image_classifier_model.keras')
+
+loss, accuracy = model.evaluate(val_generator)
+print(f'Validation Loss: {loss:.4f}')
+print(f'Validation Accuracy: {accuracy:.4f}')

model.py

@@ -0,0 +1,20 @@
+import tensorflow as tf
+from tensorflow.keras import layers, models
+from tensorflow.keras.applications import InceptionV3
+
+def create_model():
+    base_model = InceptionV3(weights='imagenet', include_top=False, input_shape=(150, 150, 3))
+    base_model.trainable = False  # Freezing the base model layers
+
+    model = models.Sequential([
+        base_model,
+        layers.GlobalAveragePooling2D(),
+        layers.Dense(512, activation='relu'),
+        layers.Dropout(0.5),
+        layers.Dense(1, activation='sigmoid')
+    ])
+
+    model.compile(optimizer=tf.keras.optimizers.Adam(learning_rate=0.0001), 
+                  loss='binary_crossentropy', 
+                  metrics=['accuracy'])
+    return model

pyproject.toml

@@ -0,0 +1,25 @@
+[tool.poetry]
+name = "xray-image-classifier"
+version = "0.1.0"
+description = ""
+authors = ["Your Name <[email protected]>"]
+readme = "README.md"
+
+[tool.poetry.dependencies]
+python = "3.11.8"
+tensorflow = "2.15.1"
+keras = "^2.15.0"
+numpy = "^1.23.5"
+pandas = "^2.2.2"
+matplotlib = "^3.9.2"
+jupyter = "^1.0.0"
+scipy = "^1.14.1"
+[tool.poetry.group.dev.dependencies]
+pytest = "^8.3.2"
+ipython = "^8.26.0"
+autopep8 = "^2.3.1"
+jupyter = "^1.0.0"
+
+[build-system]
+requires = ["poetry-core"]
+build-backend = "poetry.core.masonry.api"

train_model.py

@@ -0,0 +1,56 @@
+import os
+import numpy as np
+import tensorflow as tf
+import matplotlib.pyplot as plt
+from tensorflow.keras.preprocessing.image import ImageDataGenerator
+from model import create_model 
+
+base_dir = 'data/chest_xray'
+train_dir = os.path.join(base_dir, 'train')
+val_dir = os.path.join(base_dir, 'val')
+
+train_datagen = ImageDataGenerator(
+    rescale=1./255,
+    rotation_range=20,
+    width_shift_range=0.2,
+    height_shift_range=0.2,
+    shear_range=0.2,
+    zoom_range=0.2,
+    horizontal_flip=True,
+    fill_mode='nearest'
+)
+val_datagen = ImageDataGenerator(rescale=1./255)
+
+train_generator = train_datagen.flow_from_directory(
+    train_dir,
+    target_size=(150, 150),
+    batch_size=32,
+    class_mode='binary'
+)
+
+val_generator = val_datagen.flow_from_directory(
+    val_dir,
+    target_size=(150, 150),
+    batch_size=32,
+    class_mode='binary'
+)
+
+sample_images, _ = next(train_generator)
+for i in range(5):
+    plt.subplot(1, 5, i+1)
+    plt.imshow(sample_images[i])
+    plt.axis('off')
+plt.show()
+
+model = create_model()
+
+history = model.fit(
+    train_generator,
+    steps_per_epoch=243,
+    epochs=10,
+    validation_data=val_generator,
+    validation_steps=280,
+    callbacks=[tf.keras.callbacks.EarlyStopping(monitor='val_loss', patience=3, restore_best_weights=True)]
+)
+
+model.save('xray_image_classifier_model.keras')

xray_image_classifier_model.keras

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad1c91968bb830cc6f96327676f487e94763aaec1c83be18e6270cb47bf273fa
+size 100786904