Dear organizers, I have some questions about the prediction goal and output format. 1) I want to confirm if semantic segmentation is the goal, although "instances" annotation masks are provided. 2) I came across the "Docker Submission" page in the "How to Participate" wiki, which states that challengers are required to submit prediction results in the same format as the training data. Can you confirm if the output format will be the example provided below? - Outputs in the type segmentation task, - {input image name}_type.png - Outputs in the part segmentation task, - {input image name}_type.png - {input image name}_parts.png Thank you.

Created by Jihun Yoon nicezebra
Yes, we expect the format as you mentioned!

Questions about the prediction goal and output format page is loading…