Question about different number of objects

[Ang-Li-95]

Dear Jaesik,

I am trying to change the code to make it work for my case, which in the training data there are different numbers of objects instead of just 3, 6 or 12 as shown in the paper. I am wondering whether there are some ways to make that kind of network trainable if the input number of objects are not a constant?

Thanks, Ang