This section contains tools for the concept detection sub-task of ImageCLEF Caption. It comprises the following tools:
- imageclef-aae: a TensorFlow program for unsupervised representation learning using auto-encoders with adversarial networks.
- notebooks: A set of Jupyter notebooks for training classifiers for the ImageCLEF concept detection task.
- concept-count is a simple CLI tool which counts all concepts in the concepts file.
- merge-submission is a simple CLI tool to merge submission files for the concept detection task.
These tools were originally developed in the scope of the following paper:
Eduardo Pinho and Carlos Costa. Feature Learning with Adversarial Networks for Concept Detection in Medical Images: UA.PT Bioinformatics at ImageCLEF 2018, CLEF working notes, CEUR, 2018. Online: http://ceur-ws.org/Vol-2125/paper_139.pdf