Spaces:
Sleeping
Sleeping
File size: 3,954 Bytes
595b88e 4a6169c 595b88e 278bdbf 595b88e 4a6169c 278bdbf 4a6169c 595b88e 4a6169c 278bdbf 595b88e 4a6169c 278bdbf 595b88e 278bdbf 595b88e 278bdbf 595b88e 278bdbf 595b88e | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 | """
Gradio interface for Book Cover Analyzer - optimized for Hugging Face Spaces
"""
import gradio as gr
from analyzer.book_analyzer import BookCoverAnalyzer
from PIL import Image
import tempfile
import os
# Initialize analyzer once (loads models)
print("Loading models... This may take 30-60 seconds on first run.")
analyzer = BookCoverAnalyzer(verbose=True)
print("Models loaded successfully!")
def analyze_book_cover(image):
"""Analyze uploaded book cover image"""
if image is None:
return "Please upload an image.", "", "", "", ""
tmp_path = None
try:
# Save PIL Image to temporary file (analyzer expects file path)
with tempfile.NamedTemporaryFile(delete=False, suffix='.jpg') as tmp:
image.save(tmp.name, format='JPEG')
tmp_path = tmp.name
# Analyze
result = analyzer.analyze(tmp_path)
# Check if analysis succeeded
if not result.get('success', False):
error = result.get('error', 'Unknown error')
return f"Analysis failed: {error}", "", "", "", ""
# Extract interpretation
interp = result.get('interpretation', {})
# Format outputs
title = interp.get('inferred_title') or "Unable to determine"
authors = ", ".join(interp.get('inferred_authors', [])) or "Unable to determine"
publisher = interp.get('inferred_publisher') or "Unable to determine"
# Format detected elements
image_regions = result.get('image_regions', [])
if image_regions:
elements_text = f"**Detected {len(image_regions)} visual elements:**\n"
for elem in image_regions[:5]: # Top 5
elements_text += f"- {elem['location']}: {elem['pytorch_class']} ({elem['confidence']:.1%})\n"
else:
elements_text = "No visual elements detected (possibly plain text cover)"
# Format detected text
text_regions = result.get('text_regions', [])
if text_regions:
text_output = f"**Detected {len(text_regions)} text regions:**\n"
for region in text_regions[:10]: # Top 10
text_output += f"- {region['text']}\n"
# Format other text (reviews/quotes)
other = interp.get('other_text', [])
if other:
text_output += f"\n**Reviews/Quotes:**\n"
for text in other[:3]:
text_output += f"- {text}\n"
else:
text_output = "No text detected on cover"
return title, authors, publisher, elements_text, text_output
except Exception as e:
error_msg = f"Error analyzing image: {str(e)}"
return error_msg, "", "", "", ""
finally:
# Always clean up temp file, even if there's an error
if tmp_path and os.path.exists(tmp_path):
try:
os.unlink(tmp_path)
except:
pass # Ignore cleanup errors
# Create Gradio interface
demo = gr.Interface(
fn=analyze_book_cover,
inputs=gr.Image(type="pil", label="Upload Book Cover"),
outputs=[
gr.Textbox(label="📚 Book Title", lines=2),
gr.Textbox(label="✍️ Author(s)", lines=2),
gr.Textbox(label="🏢 Publisher", lines=1),
gr.Textbox(label="🖼️ Visual Elements (ImageNet Classification)", lines=8),
gr.Textbox(label="📝 Detected Text (EasyOCR)", lines=15)
],
title="📖 Book Cover Analyzer",
description="""
Upload a book cover image to automatically detect:
- **Book title, author, publisher** (inferred from text)
- **Visual elements** (classified using ResNet18)
- **All text regions** (extracted using EasyOCR)
Powered by OpenCV, PyTorch, and EasyOCR.
""",
examples=[
# You can add example images here if you want
],
theme="soft"
)
if __name__ == "__main__":
demo.launch(server_name="0.0.0.0", server_port=7860)
|