Tuan Duc Ngo
Tuan Duc Ngo
Hi @ayushjain1144 , I'm also facing the same problem. Have you found the solution yet?
Hi @ayushjain1144 , did you use the V1.2_aligned version or the V1.2 version of the 3D point cloud?
Hi, For processing the original mesh files from the Replica dataset, we use `trimesh`, which automatically converts polygon meshes to triangular meshes and run the superpoint segmentation from ScanNet. We...
I think the error is this `CUDA_ARCHITECTURES is empty for target "cmTC_67ee4"`. Maybe you installed the CUDA version which is not compatible with your machine's GPU.
Hi @LinLin1031, You could try these repos [https://github.com/fpthink/SPNet](https://github.com/fpthink/SPNet) and [https://github.com/loicland/superpoint_graph](https://github.com/loicland/superpoint_graph). Best.
You could check similar issues on the original repo of `spconv`: [https://github.com/traveller59/spconv/issues/406](https://github.com/traveller59/spconv/issues/406), [https://github.com/mit-han-lab/bevfusion/issues/82](https://github.com/mit-han-lab/bevfusion/issues/82). Best.
Please give me more details: which line of code and which file, or the full traceback of this error?
It means that your input point cloud does not contain any foreground instance, leading to an empty list. You may have cropped too many points in the preprocessing step. One...
I haven't tried training the model with 200k iterations without gradient accumulations. But I did train the model with 50k iterations on 8gpus with the same learning rate of 0.0005...
Could you print the whole error log?