Update app.py

app.py

@@ -1,28 +1,35 @@
 import gradio as gr
 import numpy as np
 from tensorflow.keras.models import load_model
+from tensorflow.keras.preprocessing.image import img_to_array
 from PIL import Image
+from huggingface_hub import hf_hub_download
 
-# Load pre-uploaded model
-model = load_model("xception_model.h5")
-
-def preprocess_image(image):
-    image = image.resize((299, 299)).convert("RGB")  # ✅ Fix: match Xception input shape
-    img_array = np.array(image) / 255.0
-    return np.expand_dims(img_array, axis=0)
+# Download the model from the Hugging Face model hub
+model_path = hf_hub_download(repo_id="Zeyadd-Mostaffa/cv_GP", filename="xception_model.h5")
+model = load_model(model_path)
 
+# Preprocessing and prediction
 def predict(image):
-    img = preprocess_image(image)
-    prob = model.predict(img)[0][0]
-    label = "REAL" if prob >= 0.5 else "FAKE"
-    return {"REAL": 1 - prob, "FAKE": prob}, f"Prediction: {label}"
+    # Resize image to expected shape (299x299x3 for Xception)
+    image = image.resize((299, 299))
+    image = img_to_array(image)
+    image = np.expand_dims(image, axis=0)
+    image = image / 255.0  # Normalize
+
+    prob = model.predict(image)[0][0]
+    label = "Fake" if prob > 0.5 else "Real"
+    confidence = round(float(prob if prob > 0.5 else 1 - prob), 3)
+    return f"{label} ({confidence})"
 
-demo = gr.Interface(
+# Gradio UI
+iface = gr.Interface(
     fn=predict,
     inputs=gr.Image(type="pil"),
-    outputs=[gr.Label(num_top_classes=2), gr.Text()],
+    outputs=gr.Text(),
     title="Deepfake Detection (Xception Model)"
 )
 
-demo.launch()
+iface.launch()
+