ali6parmak / visual-pdf-segmentation

Visual PDF Segmentation

A service for extracting useful information from PDF files.

---

Table of Contents

• Dependencies
  • Installing torch
  • Installing detectron2
• Getting Started
• Usage

Dependencies

• Python >= 3.11
• poppler >= 23.04.0
• torch >= 2.1.1
• detectron2

Installing torch

If your CUDA drivers properly installed, installing torch should be fairly simple. You can refer to this link: PyTorch

For previous versions of torch, refer to this link. To install torch 2.1.1 with CUDA 11.8, you can use:

pip install torch==2.1.1 torchvision==0.16.1 torchaudio==2.1.1 --index-url https://download.pytorch.org/whl/cu118

To check if torch properly installed and CUDA is available:

import torch

print(torch.cuda.is_available())

Installing detectron2

Installing detectron2 can be a little bit tricky. Before starting installing it, please ensure that your CUDA drivers are properly set up. Then, please follow these steps in order:

• Run the following commands in terminal (if you have not done it in past). If your Python version is not 3.11, you should change the version in the third command:

sudo add-apt-repository ppa:deadsnakes/ppa -y
sudo apt update
sudo apt install python3.11-dev

• Install detectron2 via:

python -m pip install git+https://github.com/facebookresearch/detectron2.git

Getting Started

Follow these steps to run the project on your local machine.

Prerequisites

Make sure you have the following installed:

• CUDA drivers (if using GPU acceleration).
• Other dependencies as listed in the Dependencies section.

---

1. Clone the repository to your local machine:

git clone https://github.com/ali6parmak/visual-pdf-segmentation.git

2. Install the required Python packages (If you have already installed the torch, consider commenting it out inside requirements.txt):

pip install -r requirements.txt

Usage

Run:

python src/segment_pdf.py path/to/pdf.pdf path/to/pdf2.pdf path/to/pdf3.pdf

How to use GPU

To use the GPU in the docker containers

1. Install the package:

   nvidia-container-toolkit

2. Restart docker service

   systemctl restart docker

3. Start the service with

   make start_gpu