Supervised Pollen Grain Detection

This repository contains the supervised implementation of pollen detection described in our paper Airborne pollen grain detection from partially labelled data utilising semi-supervised learning. For the semi-supervised part of the implementation please check out https://github.com/MilesGrey/ssl-pollen-detection.

Installation

Clone the repository and install the necessary python packages via pip, i.e. like so:

pip install -r requirements.txt

Furthermore, the datasets should be made available in a directory datasets located at the root of the project. Datasets are made available upon request.

Training

The code provides a simple command line interface built with click to train neural networks. Training can be started by running:

python lightning_training.py --experiment_name=<EXPERIMENT_NAME>

There are several options to customize the training a full list of options can be shown with command:

python lightning_training.py --help

The permitted arguments for non-self-explanatory options are detailed below.

Option	Values
--backbone	'resnet50', 'efficient_net_v2', 'mobile_net_v3'
--classification_loss_function	'cross_entropy', 'focal_loss'
--data_augmentation	'vertical_flip', 'horizontal_flip', 'rotation', 'rotation_cutoff', 'crop'
--train_dataset	'train_synthesized_2016_augsburg15', 'train_synthesized_2016_2018_augsburg15', 'train_raw_2016_2018_augsburg15'
--validation_dataset	'validation_synthesized_2016_augsburg15', 'validation_synthesized_2016_2018_augsburg15', 'validation_raw_2016_2018_augsburg15'
--test_dataset	'test_synthesized_2016_augsburg15', 'test_synthesized_2016_2018_augsburg15', 'test_raw_2016_2018_augsburg15'

The --data_augmentation option can be used repeatedly to specify multiple data augmentations.

Evaluation

Similarly to training, you can run the evaluation as follows:

python run_evaluation.py --checkpoint_path=<PATH_TO_TRAINED_MODEL> --evaluation_dataset_group=<EVALUATION_DATASET_GROUP>

Here, --evaluation_dataset_group can have the following values:

Option	Values
--evaluation_dataset_group	'evaluate_2016augsburg15', 'evaluate_2016+2018augsburg15_raw', 'evaluate_2016+2018augsburg15_synthesised'

Citation

@article{jin2023,
    title = {Airborne pollen grain detection from partially labelled data utilising semi-supervised learning},
    journal = {Science of The Total Environment},
    pages = {164295},
    year = {2023},
    issn = {0048-9697},
    doi = {https://doi.org/10.1016/j.scitotenv.2023.164295},
    url = {https://www.sciencedirect.com/science/article/pii/S0048969723029169},
    author = {Benjamin Jin and Manuel Milling and Maria Pilar Plaza and Jens O. Brunner and Claudia Traidl-Hoffmann and Björn W. Schuller and Athanasios Damialis},
    keywords = {Aerobiology, Automatic detection, Object detection, Semi-supervised learning, Deep learning, Pollen taxonomy},
}