Skip to content

Latest commit

 

History

History
122 lines (104 loc) · 6.64 KB

GETSTARTED.md

File metadata and controls

122 lines (104 loc) · 6.64 KB
Raw

🚀 Getting Started

Installing VSenseBox is very easy and straightforward. You can install from PyPI directly or use the prebuilt .whl files on GitHub releases or install from GitHub directly or build it from source on your own machine. However, in order to get it work, you need to install all the necessary dependencies or requirements for the modules you need.

⚙️ Requirements

All requirements are not strictly limited. However, some specific modules might need some special dependencies; for example, YOLO_Classic (With .weights model) runs faster using OpenCV DNN with GPU. In this case, you might need to build OpenCV from source with GPU support or use our pyppbox-opencv instead of the official opencv-contrib-python.

  • Prerequisite:

    • Python [3.9-3.12]
    • Local VSenseBox repo: git clone https://github.com/numediart/vsensebox.git

  • Before you install dependencies/requirements:

    • For Linux, recommend changing python3 to python: sudo apt install python-is-python3
    • If you prefer conda + Python [3.9-3.12]: conda create --name vsensebox_env python=3.11
    • Upgrade pip and setuptools: 
      python -m pip install --upgrade pip
pip install "setuptools>=67.2.0"
    • Recommend uninstalling the official Ultralytics: 
      pip uninstall -y ultralytics

  • Install dependencies/requirments under vsensebox/requirements/:

    • On Windows:
      • With GPU: 
        pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121
pip install -r requirements.txt
      • For CPU-only: 
        pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu
pip install -r requirements.txt
    • On Linux:
      • With GPU: 
        python -m pip install tensorflow[and-cuda] # TensorFlow GPU
pip install torch torchvision torchaudio
pip install -r requirements.txt
      • For CPU-only: 
        python -m pip install tensorflow # TensorFlow CPU
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu
pip install -r requirements.txt
    • On macOS:
      • With GPU: Not available!
      • For CPU: 
        pip install torch torchvision torchaudio
pip install -r requirements.txt

  • ATTENTION ⚠️⚠️⚠️

    • Default configurations of VSenseBox are set to use GPU, and to switch to CPU, you need to set cpu as string for the parameter device in the YAML config file; for example, line #8 in yolo_ultralytics_v8s.yaml.

💽 Setup

vsensebox is the main package and if you use the classic YOLO detectors and DeepSORT tracker, etc., you need to also install vsensebox-data-xxx.

  • Install vsensebox

    • Download and install the latest wheel in GitHub releases or install from PyPI: 
      pip install vsensebox
    • Or install directly from GitHub: 
      pip install git+https://github.com/numediart/vsensebox.git
    • Or build from source: 
      pip install setuptools wheel build PyYAML
python -m build --wheel --skip-dependency-check --no-isolation

  • Install vsensebox-data-xxx

    • Download the latest from GitHub releases and install
    • Or install the ones you need directly from the links below: 
      pip install https://github.com/numediart/vsensebox-data/releases/download/v0.0.0/vsensebox_data_yolocls-0.0.0-py3-none-any.whl
pip install https://github.com/numediart/vsensebox-data/releases/download/v0.0.1/vsensebox_data_yoloult-0.0.1-py3-none-any.whl
pip install https://github.com/numediart/vsensebox-data/releases/download/v0.0.0/vsensebox_data_deepsort-0.0.0-py3-none-any.whl

  • Let's try some basic features of VSenseBox

    • Configurator GUI can be called in Python terminal: 
      import vsensebox
vsensebox.config()
      Now you should see the Configurator GUI like in this scheenshot:
    • You can also easily reset the internal configurations by calling the reset(). THIS CAN'T BE REVERSED! ⚠️ 
      vsensebox.reset()
    • For the details of GUI functions and configurations, check Configurations page.
    • Check the Examples page for some real coding!

  • For Linux, if the GUI does not work, you might need to install these:

    sudo apt-get install '^libxcb.*-dev' libx11-xcb-dev libglu1-mesa-dev libxrender-dev libxi-dev libxkbcommon-dev libxkbcommon-x11-dev

  • For Ubuntu on WSL 2, if the GUI does not work, you need to install these:

    sudo apt-get install libgl1-mesa-glx xdg-utils libegl1

📢 FYI

1️⃣ Customized OpenCV

OpenCV is widely used in many well-known packages, but the majority of the prebuilt WHLs on the Internet including the official one on PyPI do not include GPU support. Thus, we build our custom one which includes NVIDIA CUDA & cuDNN supports for the OpenCV DNN module. In order to well distinguish from the rest, we decided to build and change the package name from opencv-contrib-python to pyppbox-opencv -> [Repo] [WHL]

2️⃣ Customized Ultralytics

Also, similar to pyppbox-opencv, our custom ultralytics is changed to vsensebox-ultralytics, but this time, the module name is still the same ultralytics and it is the main reason why the official ultralytics must be removed. Find out more why VSenseBox needs the customized vsensebox-ultralytics -> [Repo] [PyPI]