The scripts are based on the official label studio integration with fifty-one .

First start Label studio:

label-studio start 

Then run the annotation script to automatically create a label studio project and start annotating a dataset.

python annotate.py

• name: A name for the dataset.
• anno_key: A name for the labelstudio project.
• dataset_dir: The path to the repository containing the dataset

See here for more details about the labeling schema.

• label_field: a string indicating a new or existing label field to annotate

• label_type: a string indicating the type of labels to annotate. The possible label types are:

  • "classification": a single classification stored in Classification fields
  • "detections": object detections stored in Detections fields
  • "instances": instance segmentations stored in Detections fields with their mask attributes populated
  • "polylines": polylines stored in Polylines fields with their filled attributes set to False
  • "polygons": polygons stored in Polylines fields with their filled attributes set to True
  • "keypoints": keypoints stored in Keypoints fields
  • "segmentation": semantic segmentations stored in Segmentation fields

• classes: a list of strings indicating the class options for label_field or all fields in label_schema without classes specified.

To use Label studio, an API key must be provided, see here.

Run the annotation script to automatically augment a dataset using albumentation.

python annotate.py

• name: Name of the dataset to load.
• json_file: JSON file containing labels.
• image_root: The directory containing the images.
• list_augmentations_file: A file containing a list of albumentation augmmentations.
• output_folder: A folder to save the augmented images.

Inspect a dataset by running:

python augmented_fo_dataset.py

• name: Name of the dataset to load.
• labels_path: JSON file containing labels.
• data_path: The directory containing the images.

The script is based on the official Detectron2 integration with fifty-one .

Load the dataset, split it into training/validation and train a model.

python detectron_ready.py