cvlab-gui 0.0.31

Object Detection annotation tool entirely written in Python. Automatically annotate entire images using a pretrained object detection model or pseudo-classify single bbounding boxes with Self-Supervised pretrained models.

CVLAB [WIP]

CVLAB is a computer vision annotation tool written in python. pyimgui has been adopted for the graphical user interface.

My goal is to realize a simple, lightweight and standalone app that let you annotate your images, run auto-annotation experiments with your own models, import/export data collections and analyse your data distribution.

Right now the project is a work in progress. Fundamental features are available and operative, but not ready for a daily usage yet.

Above: Settings & Info section. Allows you to setup your labels and visualize stats about your data.

Install from PyPI (Windows only)

pip install cvlab-gui

Installation requirements

pyimgui

Yolo-LAB runs on the docking branch of PyImgui (not on PyPI atm). You can install the provided wheels.

windows

pip install *win_amd64.whl

linux

pip install *linux_x86_64.whl

other requirements

pip install -r requirements.txt
pip install -r yolov5/requirements.txt

Setup new project

python -m cvlab_gui.setup_project

Run CVLAB GUI

python -m cvlab_gui.main