get raw output (#1)

- get raw output (0d5e20680c409953a7f3c807bbe5f524d5e1ce22)

app.py

@@ -8,6 +8,7 @@ import cv2
 import numpy as np
 from PIL import Image
 from typing import Tuple, List
+import json
 from rfdetr.detr import RFDETRMedium
 
 # UI Element classes
@@ -76,26 +77,47 @@ def draw_detections(
 
     return img_with_boxes
 
+
+def detections_to_raw_json(detections) -> str:
+    out = []
+    for box, score, cls_id in zip(
+        detections.xyxy,
+        detections.confidence,
+        detections.class_id
+    ):
+        cid = int(cls_id)
+        out.append({
+            "class_id": cid,
+            "class_name": CLASSES[cid] if 0 <= cid < len(CLASSES) else str(cid),
+            "score": float(score),
+            "box_xyxy": [
+                float(box[0]),
+                float(box[1]),
+                float(box[2]),
+                float(box[3]),
+            ],
+        })
+    return json.dumps(out, indent=2)
+
+
 @torch.inference_mode()
 def detect_ui_elements(
     image: Image.Image,
     confidence_threshold: float,
     line_thickness: int
-) -> Tuple[Image.Image, str]:
+) -> Tuple[Image.Image, str, str]:
     """
     Detect UI elements in the uploaded image
-
     Args:
         image: Input PIL Image
         confidence_threshold: Minimum confidence score for detections
         line_thickness: Thickness of bounding box lines
-
     Returns:
         Annotated image and detection summary text
     """
     try:
         if image is None:
-            return None, "Please upload an image first."
+            return None, "Please upload an image first.", "[]"
 
         # Load model
         model = load_model()
@@ -130,20 +152,22 @@ def detect_ui_elements(
         # Create summary text
         summary_text = f"**Total detections:** {len(filtered_boxes)}"
 
-        return annotated_pil, summary_text
+        raw_json = detections_to_raw_json(detections)
+
+        return annotated_pil, summary_text, raw_json
 
     except Exception as e:
         import traceback
         error_msg = f"**Error during detection:**\n\n```\n{str(e)}\n\n{traceback.format_exc()}\n```"
         print(error_msg)  # Also print to logs
-        return None, error_msg
+        return None, error_msg, "[]"
+
 
 # Gradio interface
 with gr.Blocks(title="UI-DETR-1 UI Element Detector", theme=gr.themes.Soft()) as demo:
 
     gr.Markdown("""
     # UI-DETR-1 UI Element Detector
-
     Upload a screenshot or UI mockup to automatically detect elements.
     """)
 
@@ -185,14 +209,18 @@ with gr.Blocks(title="UI-DETR-1 UI Element Detector", theme=gr.themes.Soft()) as
 
             summary_output = gr.Markdown(label="Detection Summary")
 
+            raw_output = gr.Code(
+                label="Raw Detection",
+                language="json"
+            )
 
     # Connect button
     detect_button.click(
         fn=detect_ui_elements,
         inputs=[input_image, confidence_slider, thickness_slider],
-        outputs=[output_image, summary_output]
+        outputs=[output_image, summary_output, raw_output]
     )
 
 # Launch
 if __name__ == "__main__":
-    demo.queue().launch(share=False)
+    demo.queue().launch(share=False)