This plant science problem is important for field robotics where the mechanical detection of weed is a current challenge to be addressed to avoid the use of phytochemical products.
The first part of the challenge (from 20.11.2018 to 07.01.2019) was done on restricted registration, validate according to the procedure defined by the website https://agtech-challenge.com/comment-participer/. The challenge is already open to all audiences with an unlimited number of submissions.
The problem is a binary classification problem. Each input is a patch of an image with plants and weeds. You must classify weeds from plants. Each line represents the probabilities of class membership, which sum up to one. Preparing your submission with the starting kit is the easiest.
There are 2 phases:
You only need to submit the prediction results (no code).
The submissions are evaluated using the mse_metric metric. This metric computes the balanced accuracy (that is the average of the per class accuracies). The metric is re-scaled linearly between 0 and 1, 0 corresponding to perfect predictions and 1 to a random guess.
Notice: Score of -1 means that there is a problem in submitted files.
Submissions must be made before the end of phase 1. You are allowed to submit 6 submissions (maximum). Your last submission of phase 1 will be automatically forwarded to phase 2
Start: Jan. 17, 2019, 10 a.m.
Description: Development phase: create models and submit results (prediction) on validation and test data; feed-back are provided on the validation set only.
Start: March 14, 2019, 7 a.m.
Description: Final phase: Test data will be sent to participant one week before the starting of the final phase. Submissions from the previous phase are automatically cloned and used to compute the final score. The results on the test set will be revealed when the organizers make them available.
You must be logged in to participate in competitions.Sign In