8 Commits

Author	SHA1	Message	Date
wassname	0a6b08b83f	allow inputs with 3 channels (use_xyz)	2018-04-10 13:28:20 +08:00
wassname	96f46514d6	updated broken imports for new model modules	2018-04-10 10:51:11 +08:00
erikwijmans	0515ba5c69	Updates to data augmentation	2018-02-22 20:08:56 -05:00
erikwijmans	65a127f3d2	Some faster/better kernels. Tensors with points are now kept in (b, c, ...) format as this is easier for pytorch	2018-02-10 20:32:52 -05:00
erikwijmans	c4ddd6b099	memcpy is super slow in cuda kernels....	2018-01-29 18:22:27 -05:00
erikwijmans	5a5adc2b77	Updates	2018-01-06 12:13:52 -05:00
erikwijmans	803d7e1fc6	Updates and some refactoring	2017-12-26 19:49:52 -05:00
erikwijmans	dc4e2b0db3	Initial commit	2017-12-26 18:43:17 -05:00