📄 Document Summarizer

Easily upload a PDF or image (JPEG/PNG) and get a summary using T5 or BART transformer models. Includes a simple web GUI!

---

🏗️ Project Architecture

• Frontend: Streamlit web app for file upload, model selection, and displaying summaries.
• Backend: FastAPI server for handling file uploads, OCR, and text summarization.
• OCR Engine: Tesseract (via pytesseract) for extracting text from images and PDFs.
• Summarization Models: HuggingFace Transformers (T5 and BART) for abstractive text summarization.
• Storage: Uploaded files and generated outputs are stored in dedicated folders for traceability.

---

🧑‍💻 Technology Stack

• Python 3.8+
• FastAPI for high-performance async API
• Streamlit for rapid, beautiful web UI
• PyMuPDF (fitz) for PDF parsing
• Pillow for image processing
• pytesseract for OCR
• transformers (HuggingFace) for state-of-the-art NLP models
• torch for deep learning backend

---

🤖 Model Details

• T5 (Text-to-Text Transfer Transformer):
  • Pre-trained on a large corpus for general-purpose text summarization.
  • Handles a wide range of document types and writing styles.
• BART (Bidirectional and Auto-Regressive Transformers):
  • Excels at abstractive summarization and paraphrasing.
  • Robust to noisy or unstructured input.

Both models are loaded at startup for fast inference and can be easily swapped or extended with other HuggingFace models.

---

🔌 Extensibility

• Add More Models: Plug in any HuggingFace summarization model with minimal code changes.
• API-First: Easily integrate with other apps or workflows via REST API.
• Custom OCR: Swap Tesseract for another OCR engine if needed.
• Cloud Ready: Can be containerized and deployed to cloud platforms (Dockerfile not included by default).

---

💡 Use Cases

• Summarize research papers, reports, or contracts
• Extract and condense text from scanned documents
• Quickly review meeting notes or handwritten memos
• Build into larger document management or knowledge systems

---

🛠️ Quick Start Guide

1. Clone the Repository

git clone https://github.com/your-username/document-summarizer.git
cd document-summarizer

2. Install Python Dependencies

pip install -r requirements.txt

3. Install Tesseract OCR (Required for Image/PDF Text Extraction)

• Download for Windows:
  • UB Mannheim Tesseract Installer (recommended)
  • Install to the default location (e.g., C:\Users\YOURNAME\AppData\Local\Programs\Tesseract-OCR)
• No need to add Tesseract to PATH!
• The code is set up to use the default install location. If you install Tesseract elsewhere, update this line in app/ocr.py:

  pytesseract.pytesseract.tesseract_cmd = r"C:\\Users\\YOURNAME\\AppData\\Local\\Programs\\Tesseract-OCR\\tesseract.exe"

4. Start the Backend (API)

uvicorn app.main:app --reload

• You should see: Uvicorn running on http://127.0.0.1:8000
• Test the API at: http://127.0.0.1:8000/docs

5. Start the Web GUI (Frontend)

streamlit run app/gui.py

• Open http://localhost:8501 in your browser.
• Upload a PDF or image, select a model, and click Summarize!

---

✨ Features

• Upload PDF or image files (PNG, JPG, JPEG)
• Choose between T5 and BART summarization models
• Get a concise summary in seconds
• Clean, modern, and user-friendly interface
• Error handling and loading indicators for a smooth experience

---

🛠️ Troubleshooting

• Tesseract not found?
  • Make sure you installed Tesseract as above.
  • If you installed to a custom location, update the path in app/ocr.py as shown above.
• Can't connect to backend?
  • Make sure you started the backend (uvicorn ...) and see the correct message in the terminal.
  • The GUI and backend must both be running at the same time.
• Other issues?
  • Restart both the backend and GUI after making changes.
  • Check for typos in file paths.

---

Developed with ❤️ using FastAPI & Streamlit.

AI, NLP, DocumentSummarizer, SummarizeX, OCR, DeepLearning, FastAPI, Streamlit, HuggingFace, Python, OpenSource, Productivity, TextSummarization, PDFTools