IMDLBenCo DocumentationIMDLBenCo Documentation
  • Basic Information

    • Introduction
    • Framework Design
  • Quick Start

    • Installation
    • Dataset Preparation
    • Case One: Reproduce SoTA Papers by Training with Model Zoo
    • Case Two: Using Model Zoo with Checkpoint for Quick Testing
    • Case Three: Implementing Your Own Model with benco init
    • Case Four: Inference and Save a Dataset's Mask and Label for Observation and Subsequent Use
  • 简介
  • Datasets

    • Tampering Detection Dataset Index
    • AIGC Generated Content Dataset Index
  • Models & Papers

    • Models Implemented in BenCo
    • Other Models, Algorithms, Papers
  • English
  • 简体中文
GitHub
  • Basic Information

    • Introduction
    • Framework Design
  • Quick Start

    • Installation
    • Dataset Preparation
    • Case One: Reproduce SoTA Papers by Training with Model Zoo
    • Case Two: Using Model Zoo with Checkpoint for Quick Testing
    • Case Three: Implementing Your Own Model with benco init
    • Case Four: Inference and Save a Dataset's Mask and Label for Observation and Subsequent Use
  • 简介
  • Datasets

    • Tampering Detection Dataset Index
    • AIGC Generated Content Dataset Index
  • Models & Papers

    • Models Implemented in BenCo
    • Other Models, Algorithms, Papers
  • English
  • 简体中文
GitHub
  • Guide

    • Introduction
    • Framework Design
  • Quick Start

    • Installation
    • Dataset Preparation
    • Case One: Reproduce SoTA Papers by Training with Model Zoo
    • Case Two: Using Model Zoo with Checkpoint for Quick Testing
    • Case Three: Implementing Your Own Model with benco init
    • Case Four: Inference and Save a Dataset's Mask and Label for Observation and Subsequent Use

Installation

Please note that, unlike most code provided in papers, the Benco repository is not intended to be used via methods like git clone, as the code involves numerous components required for engineering development. The expected method of usage is through pip install, treating it as a Python library.

For Regular Users

If you only wish to use IMDL-BenCo to reproduce the paper and build your own model, the installation process is very simple. Currently, IMDL-BenCo is managed via PyPI, and you can complete the installation by running the following command:

pip install imdlbenco

You can run the following command to check If the python package of IMDLBenCo is installed properly and check for the latest version. (This repository is under rapid development and will release new version offten.)

Abbreviated Command
benco -v  
Full Command
benco --version  

If you have installed latest version, you will see the following content:

IMDLBenCo codebase version: 0.1.21
        Checking for updates...
        Local version:  0.1.21
        PyPI newest version:  0.1.21
You are using the latest version: 0.1.21.

Just like other command line interface, you can run benco -h to check for help and guidance.

For Developers Contributing to the Official Repository

If you are trying to develop new features for the IMDL-BenCo Python Library and contribute to the official repository, follow the instructions in this section. It is recommended to first uninstall any previously installed versions of IMDL-BenCo in your environment. Then, clone your forked repository of IMDL-BenCo, switch to the dev branch to get the latest "development version," and use the special command pip install -e . to complete the local installation. This will ensure that the current Python environment always executes the IMDL-BenCo library based on the scripts in this directory and automatically updates the corresponding behavior when files are updated, which is highly convenient for debugging and development.

# Uninstall any existing IMDL-BenCo library
pip uninstall imdlbenco

# Clone your forked IMDL-BenCo repository from GitHub
git clone https://github.com/your_name/IMDL-BenCo.git

# Enter the project directory
cd IMDL-BenCo

# Perform a local development installation using `pip install -e .`
pip install -e .

# Verify the installation
pip show imdlbenco

If the installation is successful, after executing pip list, you should see something like this:

Package                 Version            Editable project location
----------------------- ------------------ ------------------------------------------------------
...
IMDLBenCo               0.1.10             /mnt/data0/xiaochen/workspace/IMDLBenCo_pure/IMDLBenCo
...

The presence of a corresponding path in the Editable project location column indicates that any modifications to the Python scripts in this path will take effect directly in the Python environment without the need for reinstallation. This is very convenient for debugging.

Edit this page on GitHub
Last Updated:: 3/30/25, 1:14 PM
Contributors: Ma Xiaochen (马晓晨)
Next
Dataset Preparation