Question about the code

[harborsarah]

Dear authors,

Thanks for your valuable work. I am going through the code and I have a question regarding the following code.
https://github.com/robot-learning-freiburg/BEVCar/blob/978d09c95df0672901a402a6ed729e8936324b51/nets/segnet_transformer_lift_fuse_new_decoders.py#L1095C9-L1096C74

from these two lines, self.bev_queries already has shape (B, X*Z, C), but this bev_queries_learned repeat self.bev_queries another B times, is that correct? or if I understand things wrong.

Thanks for your answering.