search

aharley / simple_bev

A Simple Baseline for BEV Perception
MIT License
457 stars 70 forks source link

mask mems about reduce_masked_mean #29

Closed fuzzypants123 closed 1 year ago

fuzzypants123 commented 1 year ago

Thansk to the excellent work! I have a question about reduce_masked_mean. in oder to fuse every camera feature, you get "mask_mems" by mask_mems = (torch.abs(feat_mems) > 0).float(), and get in reduce_masked_mean, but in vox_util.unproject_image_to_mem function , there is a "valid_mem" fromvalid_mem = (x_valid & y_valid & z_valid).reshape(B, 1, Z, Y, X).float(), and adapt to values by values = values * valid_mem , I'm confused about this, why not using "valid_mem" from vox_util.unproject_image_to_mem ,but get a new "mask_mems" by mask_mems = (torch.abs(feat_mems) > 0).float()? seems not make sense to me.

image

image

image

fuzzypants123 commented 1 year ago

guess I understand now .. sorry for the interrupting ; )