Upload app.py

app.py

@@ -0,0 +1,65 @@
+import gradio as gr
+import tensorflow as tf
+import numpy as np
+from PIL import Image
+
+# Make sure to decorate the custom metric class with @register_keras_serializable
+@tf.keras.utils.register_keras_serializable()
+class F1Score(tf.keras.metrics.Metric):
+    def __init__(self, name='f1_score', **kwargs):
+        super().__init__(name=name, **kwargs)
+        self.precision = tf.keras.metrics.Precision()
+        self.recall = tf.keras.metrics.Recall()
+
+    def update_state(self, y_true, y_pred, sample_weight=None):
+        y_pred = tf.round(y_pred)
+        self.precision.update_state(y_true, y_pred, sample_weight)
+        self.recall.update_state(y_true, y_pred, sample_weight)
+
+    def result(self):
+        p = self.precision.result()
+        r = self.recall.result()
+        return 2 * ((p * r) / (p + r + tf.keras.backend.epsilon()))
+
+    def reset_states(self):
+        self.precision.reset_states()
+        self.recall.reset_states()
+
+# Load your TensorFlow model
+model_path = './ACC0.9322_valACC0.9148_loss0.3592_Epoch83.keras'
+model = tf.keras.models.load_model(model_path, custom_objects={'F1Score': F1Score})
+
+def predict(image):
+    # Preprocess the image to match the input shape of the model
+    img = Image.fromarray(image.astype('uint8'), 'RGB')
+    img = img.resize((224, 224))  # Update the size if different
+    img_array = np.array(img)
+    img_array = img_array / 255.0  # Normalize if your model expects normalization
+    img_array = img_array[np.newaxis, ...]  # Model expects a batch dimension
+
+    # Make prediction
+    prediction = model.predict(img_array)
+    # Calculate prediction probability
+    prediction_probability = float(prediction[0][0])
+
+    # Check if prediction is 'Normal' or 'Pneumonia'
+    if prediction_probability <= 0.5:
+        # If 'Normal', add 70% to its probability but cap it at 1.0 (100%)
+        adjusted_probability = min(prediction_probability + 0.5, 1.0)
+        pred_class = 'Normal'
+    else:
+        # If 'Pneumonia', use the original probability
+        adjusted_probability = prediction_probability
+        pred_class = 'Pneumonia'
+
+    # Return the prediction class and the (adjusted) probability
+    return {pred_class: adjusted_probability}
+
+iface = gr.Interface(fn=predict,
+                     inputs=gr.components.Image(image_mode='RGB', label="Upload X-ray Image"),
+                     outputs=gr.components.Label(num_top_classes=2),
+                     title="Pneumonia Detection",
+                     description="Upload an X-ray image to detect pneumonia.")
+
+if __name__ == "__main__":
+    iface.launch(share=True)