fix: hide labels

app.py

@@ -3,22 +3,24 @@ import torch
 from PIL import Image
 from huggingface_hub import hf_hub_download
 import numpy as np
+import cv2  # Import OpenCV
 
 # Download the model from Hugging Face Hub
 model_path = hf_hub_download(repo_id='ethandavey/yoloV5-coursework-model', filename='model.pt')
 
 # Load the YOLOv5 model
 model = torch.hub.load('ultralytics/yolov5', 'custom', path=model_path, source='github')
-model.line_thickness = 3 # Set the thickness of the bounding box lines
-model.hide_labels = True
-
+model.line_thickness = 3  # Set the thickness of the bounding box lines
 
 # Print model class names
 print("Model class names:", model.names)
-
 print("Model loaded successfully.")
 
 def detect(image):
+    # Convert PIL Image to NumPy array (OpenCV uses BGR format)
+    image_np = np.array(image)
+    image_np = image_np[:, :, ::-1].copy()  # Convert RGB to BGR
+
     # Run inference
     results = model(image)
 
@@ -29,19 +31,32 @@ def detect(image):
     if len(detections) > 0:
         print(f"Number of detections: {len(detections)}")
         for i, (*xyxy, conf, cls) in enumerate(detections):
-            # Extract box information
-            xyxy = [coord.item() for coord in xyxy]  # Coordinates
+            # Extract box coordinates and convert to integers
+            xyxy = [int(coord.item()) for coord in xyxy]  # Coordinates
             conf = conf.item()  # Confidence score
             cls = int(cls.item())  # Class index
             class_name = model.names[cls]  # Class name
             print(f"Detection {i}: Class '{class_name}', Confidence {conf:.2f}, Coordinates {xyxy}")
+
+            # Draw bounding box without label
+            cv2.rectangle(
+                image_np,
+                (xyxy[0], xyxy[1]),
+                (xyxy[2], xyxy[3]),
+                color=(0, 255, 0),
+                thickness=model.line_thickness,
+                lineType=cv2.LINE_AA,
+            )
     else:
         print("No detections were made.")
 
-    # Annotate image
-    annotated_image = np.squeeze(results.render())  # Render the detections on the image
+    # Convert BGR back to RGB
+    image_np = image_np[:, :, ::-1]
+
+    # Convert NumPy array back to PIL Image
+    annotated_image = Image.fromarray(image_np)
 
-    return Image.fromarray(annotated_image)
+    return annotated_image
 
 # Create Gradio Interface
 iface = gr.Interface(
@@ -49,7 +64,7 @@ iface = gr.Interface(
     inputs=gr.Image(type="pil"),
     outputs=gr.Image(type="pil"),
     title="YOLOv5 Object Detection",
-    description="Upload an image to detect objects using the YOLOv5 model."
+    description="Upload an image to detect objects using the YOLOv5 model. Labels are hidden in the output."
 )
 
 iface.launch()

requirements.txt

@@ -9,4 +9,5 @@ seaborn
 tqdm
 pandas
 ultralytics
-huggingface-hub
+huggingface-hub
+opencv-python