Update app.py

app.py

@@ -1,30 +1,29 @@
-import cv2
-import numpy as np 
 import gradio as gr
-from tensorflow.keras.utils import img_to_array
+import cv2
+import numpy as np
 from tensorflow.keras.models import load_model
-import os
+from tensorflow.keras.utils import img_to_array
 
-model = load_model(r'deepfake_detection_mobilenet_model.h5')
+# Load the pre-trained model
+model = load_model('deepfake_detection_mobilenet_model.h5')
 
+# Define the function for prediction
 def predict_image(img):
-    x = img_to_array(img)
+    # Resize and preprocess the input image
+    x = cv2.resize(img, (224, 224))
+    x = img_to_array(x) / 255.0  # Normalize the image
+    x = np.expand_dims(x, axis=0)  # Add batch dimension
     
-    # Resize to the expected input size of the model
-    x = cv2.resize(x, (224, 224), interpolation=cv2.INTER_AREA)
+    # Predict with the model
+    prediction = (model.predict(x) > 0.5).astype("int32")[0][0]
     
-    x /= 255.0
-    x = np.expand_dims(x, axis=0)
-
-    prediction = np.argmax(model.predict(x), axis=1)
-    
-    if prediction == 0:
-        return 'Real Image'
+    # Return result based on the prediction
+    if prediction == 1:
+        return "Fake Image"
     else:
-        return 'Fake Image'
-
-# Define the Gradio Interface with the desired title and description
+        return "Real Image"
 
+# Define the Gradio Interface
 description_html = """
 <p>Upload a face image to check if it's real or morphed with deepfake</p>
 """
@@ -33,10 +32,11 @@ custom_css = """
 div {background-color: whitesmoke;}
 """
 
+# Create the Gradio app interface
 gr.Interface(
     fn=predict_image,
-    inputs='image',
-    outputs='text',
+    inputs=gr.Image(type="numpy", label="Upload Face Image"),
+    outputs=gr.Textbox(label="Prediction"),
     title="Deepfake Image Detection",
     description=description_html,
     allow_flagging='never'