Update app.py

app.py

@@ -14,21 +14,39 @@ def normalize_image(img):
     img = tf.image.resize(img, (IMG_HEIGHT, IMG_WIDTH), method='bilinear')
     return img
 
-def predict_top_classes(img):
-     img = img.convert('RGB')
-     img_data = normalize_image(img)
-     x = np.array(img_data)
-     x = np.expand_dims(x, axis=0)
-     temp = model.predict(x)
-     
-     idx = np.argsort(np.squeeze(temp))[::-1]
-     top5_value = np.asarray([temp[0][i] for i in idx[0:5]])
-     top5_idx = idx[0:5]
-
-     return {CLASS_LABEL[i]:str(v) for i,v in zip(top5_idx,top5_value)}
-
-    
-model = tf.keras.models.load_model("Xception3.h5")
-
-interface = gr.Interface(predict_top_classes, gr.inputs.Image(type='pil'), outputs='label')
+def load_model(model_name):
+    # Load the model based on the model_name input
+    if model_name == "model1":
+        return tf.keras.models.load_model("model1.h5")
+    elif model_name == "model2":
+        return tf.keras.models.load_model("model2.h5")
+    elif model_name == "model3":
+        return tf.keras.models.load_model("model3.h5")
+    else:
+        raise ValueError("Invalid model_name")
+
+def predict_top_classes(img, model_name):
+    img = img.convert('RGB')
+    img_data = normalize_image(img)
+    x = np.array(img_data)
+    x = np.expand_dims(x, axis=0)
+    model = load_model(model_name)
+    temp = model.predict(x)
+
+    idx = np.argsort(np.squeeze(temp))[::-1]
+    top5_value = np.asarray([temp[0][i] for i in idx[0:5])
+    top5_idx = idx[0:5]
+
+    return {CLASS_LABEL[i]: str(v) for i, v in zip(top5_idx, top5_value)}
+
+interface = gr.Interface(
+    predict_top_classes,
+    [
+        gr.inputs.Image(type='pil'),
+        gr.inputs.Button(label="Model 1 (Xception)", value="model1"),
+        gr.inputs.Button(label="Model 2 (InceptionV3)", value="model2"),
+        gr.inputs.Button(label="Model 3 (InceptionResNetV2)", value="model3"),
+    ],
+    outputs='label'
+)
 interface.launch()