Update app.py

app.py

@@ -4,15 +4,19 @@ from PIL import Image
 from torchvision.transforms.functional import crop
 import gradio as gr
 
+# Global variables for models
+object_detection_model = None
+captioning_model = None
+tokenizer = None
+captioning_processor = None
+
 # Load models during initialization
 def init():
     global object_detection_model, captioning_model, tokenizer, captioning_processor
 
     # Step 1: Load the YOLOv5 model from Hugging Face
     try:
-        # Load the YOLOv5 model with trust_repo=True to skip the warning
         object_detection_model = torch.hub.load('ultralytics/yolov5', 'custom', path='yolov5/weights/best14.pt', trust_repo=True)
-        # Assuming model is locally available
         print("YOLOv5 model loaded successfully.")
     except Exception as e:
         print(f"Error loading YOLOv5 model: {e}")
@@ -36,6 +40,12 @@ def crop_objects(image, boxes):
 
 # Gradio interface function
 def process_image(image):
+    global object_detection_model, captioning_model, tokenizer, captioning_processor
+
+    # Ensure models are loaded
+    if object_detection_model is None or captioning_model is None or tokenizer is None or captioning_processor is None:
+        init()  # Call init to load models
+
     try:
         # Step 1: Perform object detection with YOLOv5
         results = object_detection_model(image)
@@ -78,7 +88,7 @@ def process_image(image):
 # Initialize models
 init()
 
-# Updated Gradio interface with new syntax
+# Gradio Interface
 interface = gr.Interface(
     fn=process_image,  # Function to run
     inputs=gr.Image(type="pil"),  # Input: Image upload
@@ -88,6 +98,5 @@ interface = gr.Interface(
     live=True
 )
 
-
 # Launch the Gradio app
 interface.launch()