no name
no name
I have tried to use this copy of code, however, I found the batch_size per GPU is limited 1. It is limited by the following codes in base_dataset.py: elif key...
Hi, thanks for the great work and code. When I try to build the code, the build.sh is fine but when I run the build.py, an error occurs: "ImportError: torch.utils.ffi...
When I do: from inplace_abn import InPlaceABN, InPlaceABNSync there is an error: /home/xx/.local/lib/python3.6/site-packages/inplace_abn-1.0.7-py3.6-linux-x86_64.egg/inplace_abn/_backend.cpython-36m-x86_64-linux-gnu.so: undefined symbol: _ZN3c105ErrorC1ENS_14SourceLocationERKSs Can anyone help me? Environment: cuda 10.0 pytorch 1.2 torchvision 0.4.0 gcc 7.4.0
Thanks for sharing the code. I have noticed that in your other repositories (for example, https://github.com/Zzh-tju/DIoU-pytorch-detectron), you implement the computeciou function as the ciou loss, which is different from the...
Thanks for your great work. I have trained the KPConv (deformable) on NPM3D, but the submission result is just around 79. As I noticed the KPFCNN on the evaluation website...
Thanks to your great work. I have called the compute_ciou function to generate the bbox loss, self.bbox_loss = compute_ciou _, bbox_loss = self.bbox_loss(bbox_pred, bbox_target, bbox_inside_weight, bbox_outside_weight, transform_weights=config.network.bbox_reg_weights) However, I found...
Hi, thank you for implementing this copy of code. I'd like to ask for the running environment for this copy of code. Also, how I adapt to other datasets?
In RandLANet.py line 69, it is written: f_decoder_i = self.decoder_blocks[j](torch.cat([f_encoder_list[-j - 2], f_interp_i], dim=1)) I'm wondering why not concatenate the f_encoder_list[-j - 1] and f_interp_i?
After reading the code, I found the input is just xyz points, while no features (colors) are included. I'm not sure it is fine?
The instance segmentation loss is going to NAN even if the lr is set to 0.