Track description

The partition of the Faces of the World dataset that will be used in Track 2 contains over 8,000 images, each displaying a single individual, labelled with the accessories that they are wearing. Accessories considered are: earrings, hats, glasses, necklaces, headbands, scarves and ties/neckties. Images has been labeled by multiple individuals, using Zooniverse as a tool.

In this first stage, we provide two ZIP files containing the images for the training and validation sets. The participants will evaluate their methods in the validation set. Then, in the second stage they will submit their models to be tested by the organisers of the track.

Example of labelled images.

Please send your submissions to:, with the subject name "SUBMISSION by [Name of your Team].

We would like for you to upload your submission as a ZIP file to CodaLab or any other service and send us a link to your file. The deadline is the 30th of Match.

Your submission should include: your model, the code necessary to train it, any external dataset you may have used and a fact sheet.

Thank you all and good luck!


There are no news registered in