Upload 4 files

app.py

@@ -6,14 +6,14 @@ from PIL import Image
 from huggingface_hub import hf_hub_download
 import os
 
-# Configuration
-MODEL_REPO = "Ahmedhassan54/Image-Classification"  # Replace with your HF username and repo
+
+MODEL_REPO = "Ahmedhassan54/Image-Classification" 
 MODEL_FILE = "best_model.h5"
 
-# Download model from Hugging Face Hub
+
 def load_model_from_hf():
     try:
-        # Check if model exists locally first
+       
         if not os.path.exists(MODEL_FILE):
             print("Downloading model from Hugging Face Hub...")
             model_path = hf_hub_download(
@@ -21,10 +21,10 @@ def load_model_from_hf():
                 filename=MODEL_FILE,
                 cache_dir="."
             )
-            # Copy to current directory for easier access
+          
             os.system(f"cp {model_path} {MODEL_FILE}")
         
-        # Load the model
+        
         model = tf.keras.models.load_model(MODEL_FILE)
         print("Model loaded successfully!")
         return model
@@ -32,22 +32,21 @@ def load_model_from_hf():
         print(f"Error loading model: {str(e)}")
         raise
 
-# Load the model when the app starts
+
 model = load_model_from_hf()
 
-# Image classification function
+
 def classify_image(image):
     try:
-        # Preprocess the image
-        image = image.resize((150, 150))  # Match model's expected input size
-        image_array = np.array(image) / 255.0  # Normalize
-        image_array = np.expand_dims(image_array, axis=0)  # Add batch dimension
+       
+        image = image.resize((150, 150)) 
+        image_array = np.array(image) / 255.0  
+        image_array = np.expand_dims(image_array, axis=0) 
         
-        # Make prediction
+      
         prediction = model.predict(image_array)
         confidence = float(prediction[0][0])
         
-        # Format results
         if confidence > 0.5:
             return {
                 "Dog": confidence * 100,
@@ -61,7 +60,7 @@ def classify_image(image):
     except Exception as e:
         return f"Error processing image: {str(e)}"
 
-# Gradio interface
+
 demo = gr.Interface(
     fn=classify_image,
     inputs=gr.Image(type="pil", label="Upload Image"),
@@ -75,6 +74,6 @@ demo = gr.Interface(
     allow_flagging="never"
 )
 
-# Launch the app
+
 if __name__ == "__main__":
     demo.launch(debug=True, server_port=7860)

model.py

@@ -1,11 +1,6 @@
-# Image Classification for Cat vs Dog Dataset (Fixed Version)
-# Run in Google Colab with GPU
 
-## 1. Setup Environment
-!nvidia-smi
-!pip install tensorboard-plugin-profile
 
-## 2. Import Libraries
+
 import tensorflow as tf
 from tensorflow.keras import layers, models, callbacks
 from tensorflow.keras.preprocessing.image import ImageDataGenerator
@@ -22,49 +17,40 @@ from pathlib import Path
 
 print("TensorFlow version:", tf.__version__)
 
-## 3. Upload and Reorganize Your Dataset
-# Upload your zip file
+
 uploaded = files.upload()
 zip_filename = list(uploaded.keys())[0]
 
-# Extract the zip file
 with zipfile.ZipFile(zip_filename, 'r') as zip_ref:
     zip_ref.extractall('extracted_dataset')
 
-# Verify extraction
-!ls extracted_dataset
 
-# Your dataset has images directly in custom_dataset/train/ (not in cat/dog subfolders)
-# We need to reorganize them into proper class folders
+
+
 def organize_dataset(input_dir, output_dir):
-    # Create class directories
+ 
     os.makedirs(os.path.join(output_dir, 'cat'), exist_ok=True)
     os.makedirs(os.path.join(output_dir, 'dog'), exist_ok=True)
     
-    # Move cat images
+   
     for file in Path(input_dir).glob('cat.*.jpg'):
         move(str(file), os.path.join(output_dir, 'cat', file.name))
     
-    # Move dog images
+    
     for file in Path(input_dir).glob('dog.*.jpg'):
         move(str(file), os.path.join(output_dir, 'dog', file.name))
 
-# Reorganize the dataset
+
 input_path = 'extracted_dataset/custom_dataset/train'
 output_path = 'organized_dataset/train'
 organize_dataset(input_path, output_path)
 
-# Verify the new structure
-!ls organized_dataset/train
-!ls organized_dataset/train/cat | head -5
-!ls organized_dataset/train/dog | head -5
 
-## 4. Create Data Generators
-# Parameters
+
 IMG_SIZE = (150, 150)
 BATCH_SIZE = 32
 
-# Data generators with augmentation
+
 train_datagen = ImageDataGenerator(
     rescale=1./255,
     rotation_range=20,
@@ -73,10 +59,10 @@ train_datagen = ImageDataGenerator(
     shear_range=0.2,
     zoom_range=0.2,
     horizontal_flip=True,
-    validation_split=0.2  # 20% for validation
+    validation_split=0.2  
 )
 
-# Training generator
+
 train_generator = train_datagen.flow_from_directory(
     'organized_dataset/train',
     target_size=IMG_SIZE,
@@ -86,7 +72,7 @@ train_generator = train_datagen.flow_from_directory(
     shuffle=True
 )
 
-# Validation generator
+
 validation_generator = train_datagen.flow_from_directory(
     'organized_dataset/train',
     target_size=IMG_SIZE,
@@ -96,13 +82,12 @@ validation_generator = train_datagen.flow_from_directory(
     shuffle=True
 )
 
-# Get class names
 class_names = list(train_generator.class_indices.keys())
 print("\nDetected classes:", class_names)
 print("Number of training samples:", train_generator.samples)
 print("Number of validation samples:", validation_generator.samples)
 
-# Visualize samples
+
 plt.figure(figsize=(12, 9))
 for i in range(9):
     img, label = next(train_generator)
@@ -113,7 +98,7 @@ for i in range(9):
 plt.suptitle("Sample Training Images")
 plt.show()
 
-## 5. Build Model
+
 def build_model(input_shape):
     model = models.Sequential([
         layers.Conv2D(32, (3,3), activation='relu', input_shape=input_shape),
@@ -141,7 +126,7 @@ def build_model(input_shape):
 model = build_model(input_shape=(IMG_SIZE[0], IMG_SIZE[1], 3))
 model.summary()
 
-## 6. Train Model
+
 log_dir = "logs/fit/" + datetime.datetime.now().strftime("%Y%m%d-%H%M%S")
 
 callbacks = [
@@ -160,8 +145,7 @@ history = model.fit(
     callbacks=callbacks
 )
 
-## 7. Evaluate Model
-# Plot training history
+
 plt.figure(figsize=(12, 4))
 plt.subplot(1, 2, 1)
 plt.plot(history.history['accuracy'], label='Train')
@@ -176,10 +160,10 @@ plt.title('Loss')
 plt.legend()
 plt.show()
 
-## 8. Save Model
+
 model.save('cat_dog_classifier.h5')
 
-# Convert to TFLite
+
 converter = tf.lite.TFLiteConverter.from_keras_model(model)
 tflite_model = converter.convert()
 with open('cat_dog.tflite', 'wb') as f: