2022 ACRE Cascade Campaign

Stage 1 “Development” of the ACRE 1st Cascade Campaign is already started.

Participants are asked to develop a model to perform semantic segmentation of RGB images by training their models on the 2019 dataset.

Within the Cascade Campaign, participants are provided with the real images of maize and bean crops (plus weeds) collected by robots during ACRE’s field evaluation campaigns. Participants will also receive ground truth in the form of images labeled by human experts.
Participants are asked to use their systems to detect and mark crops and weeds in unlabeled images, which ACRE will then compare with their ground truth.

Learn more about the campaign here.

Join the ACRE Cascade Campaign on Codalab.

Learn about ACRE and its Cascade Campaigns from the 3rd ACRE Workshop recording or from our Evaluation Plan