Code for "Dense Tracking and Mapping with Voxel-based Neural Implicit Representation", ISMAR 2022
266
stars
24
forks
source link
I have some trouble running the code. I hope you can help me.Producer process has been terminated before all shared CUDA tensors released. See Note [Sharing CUDA tensors] #24
initializing first_frame: 0
initializing the first frame ...
mapping process started!
insert keyframe
tracking process started! *
tracking frame: 52%|█████████████▉ | 51/99 [00:32<00:34, 1.38it/s]insert keyframe
** current num kfs: 2 **
tracking frame: 100%|███████████████████████████| 99/99 [01:07<00:00, 1.47it/s]
========== stop_mapping set ==========
* tracking process died *
[W CudaIPCTypes.cpp:15] Producer process has been terminated before all shared CUDA tensors released. See Note [Sharing CUDA tensors]
** post-processing 0 steps **
* extracting final mesh *
/home/xyz/anaconda3/envs/Vox/lib/python3.9/site-packages/torch/functional.py:445: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at /opt/conda/conda-bld/pytorch_1634272204863/work/aten/src/ATen/native/TensorShape.cpp:2157.)
return _VF.meshgrid(tensors, kwargs) # type: ignore[attr-defined]
**** get color from network **
/home/xyz/Github/Vox-Fusion/src/utils/mesh_util.py:110: UserWarning: floordiv is deprecated, and its behavior will change in a future version of pytorch. It currently rounds toward 0 (like the 'trunc' function NOT 'floor'). This results in incorrect rounding for negative values. To keep the current behavior, use torch.div(a, b, rounding_mode='trunc'), or for actual floor division, use torch.div(a, b, rounding_mode='floor').
voxel_pos = points // self.voxel_size
tensor(1457)
* mapping process died ***
[W CudaIPCTypes.cpp:15] Producer process has been terminated before all shared CUDA tensors released. See Note [Sharing CUDA tensors]
/home/xyz/anaconda3/envs/Vox/lib/python3.9/multiprocessing/resource_tracker.py:216: UserWarning: resource_tracker: There appear to be 3 leaked semaphore objects to clean up at shutdown
warnings.warn('resource_tracker: There appear to be %d '
initializing first_frame: 0 initializing the first frame ... mapping process started! insert keyframe tracking process started! * tracking frame: 52%|█████████████▉ | 51/99 [00:32<00:34, 1.38it/s]insert keyframe ** current num kfs: 2 ** tracking frame: 100%|███████████████████████████| 99/99 [01:07<00:00, 1.47it/s] ========== stop_mapping set ========== * tracking process died * [W CudaIPCTypes.cpp:15] Producer process has been terminated before all shared CUDA tensors released. See Note [Sharing CUDA tensors] ** post-processing 0 steps ** * extracting final mesh * /home/xyz/anaconda3/envs/Vox/lib/python3.9/site-packages/torch/functional.py:445: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at /opt/conda/conda-bld/pytorch_1634272204863/work/aten/src/ATen/native/TensorShape.cpp:2157.) return _VF.meshgrid(tensors, kwargs) # type: ignore[attr-defined] **** get color from network ** /home/xyz/Github/Vox-Fusion/src/utils/mesh_util.py:110: UserWarning: floordiv is deprecated, and its behavior will change in a future version of pytorch. It currently rounds toward 0 (like the 'trunc' function NOT 'floor'). This results in incorrect rounding for negative values. To keep the current behavior, use torch.div(a, b, rounding_mode='trunc'), or for actual floor division, use torch.div(a, b, rounding_mode='floor'). voxel_pos = points // self.voxel_size tensor(1457) * mapping process died *** [W CudaIPCTypes.cpp:15] Producer process has been terminated before all shared CUDA tensors released. See Note [Sharing CUDA tensors] /home/xyz/anaconda3/envs/Vox/lib/python3.9/multiprocessing/resource_tracker.py:216: UserWarning: resource_tracker: There appear to be 3 leaked semaphore objects to clean up at shutdown warnings.warn('resource_tracker: There appear to be %d '