Upload app_hf_spaces.py with huggingface_hub

app_hf_spaces.py

@@ -0,0 +1,151 @@
+#!/usr/bin/env python3
+"""
+Hugging Face Spaces optimized app with Gradio UI
+Initializes models on startup, provides both UI and API
+"""
+
+import os
+import gc
+import json
+from pathlib import Path
+import gradio as gr
+import numpy as np
+from PIL import Image
+import cv2
+
+# Configure OmniParser
+os.environ["OMP_NUM_THREADS"] = "4"
+
+from config import get_omniparser_config
+from ui_element_analyzer import UIElementAnalyzer
+
+# Global analyzer (initialized once)
+analyzer = None
+
+def initialize_models():
+    """Initialize OmniParser models on startup"""
+    global analyzer
+    print("[INIT] Loading OmniParser models...")
+    config = get_omniparser_config()
+    analyzer = UIElementAnalyzer(config)
+    print("[INIT] Models loaded successfully")
+    gc.collect()
+
+def analyze_ui_image(image: Image.Image) -> tuple:
+    """
+    Analyze UI screenshot and return results
+    
+    Args:
+        image: PIL Image of UI screenshot
+        
+    Returns:
+        (visualization_image, json_str, csv_str, status_str)
+    """
+    if analyzer is None:
+        return None, "", "", "❌ Models not initialized. Please refresh page."
+    
+    try:
+        print("[PROCESSING] Analyzing image...")
+        
+        # Convert PIL to numpy
+        img_np = np.array(image)
+        if len(img_np.shape) == 2:
+            img_np = cv2.cvtColor(img_np, cv2.COLOR_GRAY2BGR)
+        
+        # Analyze
+        results = analyzer.analyze(img_np)
+        
+        # Load visualization
+        viz_path = results.get("visualization_path")
+        if viz_path and Path(viz_path).exists():
+            viz_image = Image.open(viz_path)
+        else:
+            viz_image = image
+        
+        # Format JSON
+        coords = results.get("coordinates", [])
+        json_str = json.dumps(coords, indent=2)
+        
+        # Format CSV
+        csv_lines = ["element_id,x,y,confidence,ocr_text"]
+        for elem in coords:
+            csv_lines.append(
+                f"{elem['element_id']},{elem['x']},{elem['y']},"
+                f"{elem['confidence']},{elem['ocr_text']}"
+            )
+        csv_str = "\n".join(csv_lines)
+        
+        status = f"✅ Detected {len(coords)} UI elements"
+        
+        return viz_image, json_str, csv_str, status
+        
+    except Exception as e:
+        print(f"[ERROR] {str(e)}")
+        return image, "", "", f"❌ Error: {str(e)}"
+
+
+# Initialize on startup
+print("[HF SPACES] Starting application...")
+initialize_models()
+
+# Create Gradio interface
+with gr.Blocks(title="OmniParser UI Detector", theme=gr.themes.Soft()) as demo:
+    gr.Markdown("# 🎯 OmniParser UI Element Detector")
+    gr.Markdown("Upload a UI screenshot to detect and locate all UI elements (buttons, text, icons, etc.)")
+    
+    with gr.Row():
+        with gr.Column(scale=1):
+            gr.Markdown("### 📤 Upload Image")
+            image_input = gr.Image(type="pil", label="UI Screenshot")
+            analyze_btn = gr.Button("🔍 Analyze", variant="primary", size="lg")
+            
+        with gr.Column(scale=1):
+            gr.Markdown("### 📊 Results")
+            viz_output = gr.Image(label="Detection Visualization")
+            status_output = gr.Textbox(label="Status", interactive=False)
+    
+    with gr.Row():
+        with gr.Column():
+            gr.Markdown("### 📋 Coordinates (JSON)")
+            json_output = gr.Textbox(
+                label="JSON Output", 
+                lines=10, 
+                max_lines=30,
+                interactive=False
+            )
+    
+    with gr.Row():
+        with gr.Column():
+            gr.Markdown("### 📈 Coordinates (CSV)")
+            csv_output = gr.Textbox(
+                label="CSV Output",
+                lines=10,
+                max_lines=30,
+                interactive=False
+            )
+    
+    # Set click handler
+    analyze_btn.click(
+        analyze_ui_image,
+        inputs=[image_input],
+        outputs=[viz_output, json_output, csv_output, status_output]
+    )
+    
+    gr.Markdown("""
+    ### ℹ️ How it works:
+    1. **Upload** a screenshot of any UI/app
+    2. **Click** "Analyze" button
+    3. **View** detected elements with coordinates
+    4. **Export** results as JSON or CSV
+    
+    Detects: Buttons, text fields, icons, dropdowns, etc.
+    """)
+
+if __name__ == "__main__":
+    print("[LAUNCH] Starting Gradio server on port 7860...")
+    demo.launch(
+        server_name="0.0.0.0",
+        server_port=7860,
+        share=False,
+        show_error=True
+    )