To submit the predicted results for evaluation, result reproduction, and ranking, participants need to send an email that includes all of the following items to leafinspect@gmail.com:
A link to their GitHub repository containing all the codes to reproduce the proposed algorithm(s).
Clear instructions to reproduce the computing environment and run their model in it. This should take the shape of a script that only requires us to change the path to the testing data to be executed.
Participants must provide a description of the methods used to produce the results submitted.
The submission format for the tasks are outlined below:
Leaf Instance Segmentation Task
A zip file containing results (predicted label images) for the test data (Figure 1).
In predicted label image, the labels are assumed to be consecutive numbers, with lowest label as background (same for ground truth labeled images). This is necessary to avoid the trivial solution (i.e. finding only background, gives excellent results).
Leaf Counting Task
A csv file listing test image names and number of leaves.
Figure 1: Sample Images with Annotated Labeled Images