minor changes

model/pothole_model.py

@@ -23,13 +23,11 @@
 #     return predicted_class
 
 
-
-from keras.models import load_model
+import tensorflow as tf
 from PIL import Image, ImageOps
 import numpy as np
 import requests
-import tempfile
-import os
+from io import BytesIO
 
 def load_image_model(image):
     # Disable scientific notation for clarity
@@ -39,14 +37,10 @@ def load_image_model(image):
     model_url = "https://huggingface.co/spaces/Soham0708/pothole_detect/blob/main/keras_model.h5"
     response = requests.get(model_url)
     response.raise_for_status()  # Raise an exception if the download fails
-    
-    # Save the model to a temporary file
-    with tempfile.NamedTemporaryFile(suffix=".h5", delete=False) as tmp_file:
-        tmp_file.write(response.content)
-        tmp_file_path = tmp_file.name
+    model_data = BytesIO(response.content)
 
-    # Load the model from the temporary file
-    model = load_model(tmp_file_path)
+    # Load the model from the in-memory bytes
+    model = tf.keras.models.load_model(model_data)
 
     # Load the labels
     class_names = open("labels.txt", "r").readlines()
@@ -57,7 +51,7 @@ def load_image_model(image):
     data = np.ndarray(shape=(1, 224, 224, 3), dtype=np.float32)
 
     # Replace this with the path to your image
-    image = Image.open(by(image)).convert("RGB")
+    image = Image.open(image).convert("RGB")
 
     # resizing the image to be at least 224x224 and then cropping from the center
     size = (224, 224)
@@ -80,7 +74,4 @@ def load_image_model(image):
 
     # Print prediction and confidence score
     print("Class:", class_name[2:], end="")
-    print("Confidence Score:", confidence_score)
-
-    # Clean up temporary file
-    os.remove(tmp_file_path)
+    print("Confidence Score:", confidence_score)