2 May 2018

The Fungi Classification Challenge


Image classification of objects has advanced greatly over the past few years. The success of computer vision is dependent on high quality datasets to train and test new algorithms. Develop the world’s best algorithm to identify fungi.

In the FUNGI CLASSIFICATION CHALLENGE, you get the chance to build algorithms based on a dataset from a carefully curated database containing over 100,000 fungi images.

The dataset holds 1,394 wild mushrooms species, with 85,578 training images and 4,182 validation images. The testing set contains 9,758 images.

The data and images are collected by citizen scientists and validated by leading fungal experts on the Danish Svampe Atlas-platform.

Submission Deadline is 4 June 2018. Training data, annotations, and links to pre-trained models are available here.

You are invited to take part in the competition on Kaggle. Winners will be announced at the Fifth Workshop on Fine-Grained Visual Categorization co-organized by BYU, Caltech, Cornell, Google, Microsoft AI for Earth, MIT, and UMass at CVPR 2018.

(Top submissions for the competition will be invited to give talks at the workshop. Attending the workshop is not required to participate in the competition, however only teams that are attending the workshop will be considered to present their work).