Hi! Could I ask why do you adopt two seperated optimizers for both segmentation network, instead of using only one optimizer to include all the trainable parameters? Are there any difference between two kinds of settings? Looking forward to your reply. Thanks!
Hi! Could I ask why do you adopt two seperated optimizers for both segmentation network, instead of using only one optimizer to include all the trainable parameters? Are there any difference between two kinds of settings? Looking forward to your reply. Thanks!