Understanding the how the visualization process works on instance segmentation

[ptang0917]

Hello,
I just want to say I am a huge fan with this framework and the work put in behind this as this has proven very useful to my work. I do have some question though. During the visualization procedure for instance segmentation, I am curious to know which object is being classified first. The screenshot below is the output we get from visualizing one of the s3dis sample dataset from the testing directory. What I'm trying to figure out is what the numbers (circled in red) represents? So far I know it corresponds to the object class list for s3dis (s3dis has 13 class objects)