zhangcheng828

icon company Hangzhou Dianzi University

Results 2 comments of zhangcheng828
trafficstars

Do you have the data loader on modelnet 40

Hi, please refer to DGCNN for loading ModelNet40 data.

Why the loss of monoflex rises to hundreds after dozens of epochs?

The dim-aware loss is particularly sensitive to abnormal data in the KITTI dataset, so I suggest using L1 loss instead of it.