Update app.py

app.py

@@ -1,51 +1,32 @@
-import os
-import numpy as np
-import matplotlib.pyplot as plt
+import gradio as gr
 import tensorflow as tf
-from tensorflow.keras.preprocessing.image import ImageDataGenerator
+from tensorflow.keras.preprocessing.image import load_img, img_to_array
 from tensorflow.keras.models import load_model
 
-# Suppress TensorFlow logging and warnings
-os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'  # Suppress TensorFlow logging (1 = INFO, 2 = WARNING)
-tf.get_logger().setLevel('ERROR')
-tf.autograph.set_verbosity(2)
-
-# Function to load the pre-trained model
-def load_pretrained_model(model_path):
-    return load_model(model_path)
-
 # Load the pre-trained model
-model_path = 'pokemon_classifier_model.keras'  # Adjust the path if your model is located elsewhere
-model = load_pretrained_model(model_path)
+model = load_model('pokemon_classifier_model.keras')
+classes = ['Doduo', 'Geodude', 'Zubat']  # List of classes
 
-# Define the Pokémon classes to be classified
-classes = ['Doduo', 'Geodude', 'Zubat']
+def classify_image(image):
+    """Function to classify the image using the pre-trained model."""
+    image = image.resize((150, 150))  # Resize image to match model's expected input
+    image_array = img_to_array(image)
+    image_array = image_array.reshape((1, 150, 150, 3))  # Reshape for model
+    image_array /= 255.0  # Normalize the image
+    
+    prediction = model.predict(image_array)
+    predicted_class = classes[np.argmax(prediction)]
+    confidence = np.max(prediction)
+    return predicted_class, f"{confidence * 100:.2f}% Confidence"
 
-# Data Augmentation and Data Generators for the predictions
-datagen = ImageDataGenerator(rescale=1./255)
-generator = datagen.flow_from_directory(
-    'pokemon_dataset',  # Adjust this path to the directory containing your data
-    target_size=(150, 150),
-    batch_size=1,  # For prediction, typically we use batch size of 1
-    classes=classes,
-    class_mode='categorical',
-    shuffle=False)
+# Create a Gradio interface
+iface = gr.Interface(
+    classify_image,
+    inputs=gr.inputs.Image(shape=(150, 150)),
+    outputs=[gr.outputs.Label(num_top_classes=3), gr.outputs.Text()],
+    title="Pokémon Image Classifier",
+    description="Upload an image of a Pokémon to classify!"
+)
 
-# Predict and visualize results
-for images, labels in generator:
-    # Use only the first 5 images for visualization
-    for i in range(5):
-        img = images[i]
-        label = labels[i]
-        plt.imshow(img)
-        # Create prediction
-        img_array = np.expand_dims(img, axis=0)  # Model expects a batch
-        predictions = model.predict(img_array)
-        predicted_class_index = np.argmax(predictions, axis=1)[0]
-        predicted_class_name = classes[predicted_class_index]
-        # Determine actual class
-        true_class_index = np.argmax(label)
-        true_class_name = classes[true_class_index]
-        plt.title(f'Predicted: {predicted_class_name}, True: {true_class_name}')
-        plt.show()
-    break  # Use only the first batch of images
+# Launch the Gradio app
+iface.launch()