again

app.py

@@ -14,24 +14,41 @@ model = fasterrcnn_resnet50_fpn_v2(weights=weights, box_score_thresh=0.9)
 #path = Path("object_detection_model.pth")
 #model = torch.load(path)
 
-def predict(inp1):
-  model.eval()
+from PIL import Image,ImageDraw
+from torchvision.io.image import read_image
+from torchvision.models.detection import fasterrcnn_resnet50_fpn_v2, FasterRCNN_ResNet50_FPN_V2_Weights
+from torchvision.utils import draw_bounding_boxes
+from torchvision import transforms
+from torchvision.transforms.functional import to_pil_image,to_tensor
+from pathlib import Path
+import torch
 
-  # Step 2: Initialize the inference transforms
-  preprocess = weights.transforms()
+# Step 1: Initiali1ze model with the best available weights
+weights = FasterRCNN_ResNet50_FPN_V2_Weights.DEFAULT
+model = fasterrcnn_resnet50_fpn_v2(weights=weights, box_score_thresh=0.9)
+preprocess = weights.transforms()
+
+def predict1(inp1):
+  model.eval()
 
-  # Step 3: Apply inference preprocessing transforms
   batch = [preprocess(inp1)]
 
   # Step 4: Use the model and visualize the prediction
   prediction = model(batch)[0]
   labels = [weights.meta["categories"][i] for i in prediction["labels"]]
-  box = draw_bounding_boxes(inp1, boxes=prediction["boxes"],
-                            labels=labels,
-                            colors="red",
-                            width=4, font_size=30)
-  im = to_pil_image(box.detach())
-  return labels[0],im
+  
+  draw = ImageDraw.Draw(inp1)
+  z = prediction["boxes"]
+
+  # Define the bounding box coordinates and size (adjust these values as needed)
+  outline_color = "red"
+  line_width = 2
+
+  # Draw the bounding box as a rectangle
+  x,y,l,k = z[0][0].item(), z[0][1].item(), z[0][2].item(), z[0][3].item()
+  draw.rectangle([x, y,l,k], outline=outline_color, width=line_width)
+
+  return labels[0],inp1
 
 
 # Get example filepaths in a list of lists