Skip to content

Latest commit

 

History

History

annotations_to_biigle

README.md

Annotations_to_biigle

  • Import pascalVOC annotations into Biigle.
  • Automatic feature detection on Biigle images using yoloV5 models and annotations import back in Biigle via pascalVOC

Installation

The script requires pandas, pytorch, seaborn, pillow, tqdm, opencv, yaml, pascal-voc-writer.

conda create --name annotations_to_biigle -y
conda activate annotations_to_biigle
conda install requests -y
conda install pytorch torchvision torchaudio cpuonly -c pytorch -y
conda install -c conda-forge pandas seaborn tqdm opencv -y
pip install pascal_voc_writer pyyaml

Alternatively, a conda environment can be installed through the environment.yml file.

Example yoloV5 detection model can be found on: https://zenodo.org/record/5539915

Usage

See the annotations_to_biigle file.

Name Description
email Your email address.
token Your Biigle token.
label_tree_id Your label tree id. This label tree MUST contain every classes possibly detected or annotated. Otherwise, a new label will automatically be added
volume_id The volume id where your images are located.
path_model Path to your yoloV5 model.
path_classes Path to a file (.txt or .names) with ALL your classes (one class per row).
path_data Path where to store downloaded images, or where image are located.
confidence Confidence detection level.
export_biigle Boolean, export to Biigle ?

Main functions

  • utils_pascalVOC.download_images : Download all images from the volume id
  • detect_yoloV5.model_inference : Do inference and export to biigle
  • export_to_biigle.pascalVOC_to_biigle : Export annotations from pascalVOC to Biigle