Update app.py

app.py

@@ -1,36 +1,50 @@
 import gradio as gr
-import numpy as np
 import tensorflow as tf
-from tensorflow.keras.models import load_model
-from tensorflow.keras.preprocessing import image
+import numpy as np
+from keras import backend as K
+from tensorflow.keras.preprocessing.image import img_to_array
+from PIL import Image
+
+# --- Custom Metrics ---
+def recall(y_true, y_pred):
+    true_positives = K.sum(K.round(K.clip(y_true * y_pred, 0, 1)))
+    possible_positives = K.sum(K.round(K.clip(y_true, 0, 1)))
+    return true_positives / (possible_positives + K.epsilon())
 
-# Load the trained model
-model = load_model("pneumonia_cnn_model.h5", custom_objects={
-    'f1': lambda y_true, y_pred: 2*((precision(y_true, y_pred)*recall(y_true, y_pred)) / (precision(y_true, y_pred)+recall(y_true, y_pred)+tf.keras.backend.epsilon())),
-    'precision': precision,
-    'recall': recall
-})
+def precision(y_true, y_pred):
+    true_positives = K.sum(K.round(K.clip(y_true * y_pred, 0, 1)))
+    predicted_positives = K.sum(K.round(K.clip(y_pred, 0, 1)))
+    return true_positives / (predicted_positives + K.epsilon())
 
-# Preprocessing function
+def f1(y_true, y_pred):
+    prec = precision(y_true, y_pred)
+    rec = recall(y_true, y_pred)
+    return 2 * ((prec * rec) / (prec + rec + K.epsilon()))
+
+# --- Load the Model ---
+model = tf.keras.models.load_model(
+    "pneumonia_cnn_model.h5",
+    custom_objects={'precision': precision, 'recall': recall, 'f1': f1}
+)
+
+# --- Prediction Function ---
 def predict_pneumonia(img):
-    img = img.convert('L')  # convert to grayscale
+    img = img.convert('L')  # Convert to grayscale
     img = img.resize((299, 299))
-    img_array = image.img_to_array(img)
-    img_array = img_array / 255.0
-    img_array = np.expand_dims(img_array, axis=0)  # batch dimension
+    img_array = img_to_array(img) / 255.0
+    img_array = np.expand_dims(img_array, axis=0)  # Shape: (1, 299, 299, 1)
     prediction = model.predict(img_array)[0][0]
+    label = "PNEUMONIA" if prediction >= 0.5 else "NORMAL"
+    confidence = prediction if prediction >= 0.5 else 1 - prediction
+    return f"{label} ({confidence * 100:.2f}% confidence)"
 
-    if prediction >= 0.5:
-        return "PNEUMONIA 🫁 (Confidence: {:.2f}%)".format(prediction * 100)
-    else:
-        return "NORMAL ✅ (Confidence: {:.2f}%)".format((1 - prediction) * 100)
-
-# Gradio interface
-interface = gr.Interface(fn=predict_pneumonia,
-                         inputs=gr.Image(type="pil"),
-                         outputs="text",
-                         title="Pneumonia Detection from Chest X-rays",
-                         description="Upload a chest X-ray image to detect Pneumonia using a CNN model.")
+# --- Gradio Interface ---
+interface = gr.Interface(
+    fn=predict_pneumonia,
+    inputs=gr.Image(type="pil"),
+    outputs=gr.Text(),
+    title="Pneumonia Detection from Chest X-ray",
+    description="Upload a chest X-ray image to detect Pneumonia using a CNN model trained on grayscale 299x299 images."
+)
 
-if __name__ == "__main__":
-    interface.launch()
+interface.launch()