-
-
-
[Project](https://alexzhou907.github.io/pvd) | [Paper](https://arxiv.org/abs/2104.03670)
@@ -11,25 +9,21 @@ Implementation of Shape Generation and Completion Through Point-Voxel Diffusion
## Requirements:
-Make sure the following environments are installed.
-
-```
-python==3.6
-pytorch==1.4.0
-torchvision==0.5.0
-cudatoolkit==10.1
-matplotlib==2.2.5
-tqdm==4.32.1
-open3d==0.9.0
-trimesh=3.7.12
-scipy==1.5.1
+Install Python environment
+```bash
+module load conda
+module load artifactory
+mamba env create --file env.yml
```
Install PyTorchEMD by
-```
+```bash
cd metrics/PyTorchEMD
+module load gcc
+module load mpfr
+conda activate PVD
python setup.py install
-cp build/**/emd_cuda.cpython-36m-x86_64-linux-gnu.so .
+cp build/**/emd_cuda.cpython-*-x86_64-linux-gnu.so .
```
The code was tested on Unbuntu with Titan RTX.
@@ -68,37 +62,29 @@ $ python train_generation.py --category car|chair|airplane --model MODEL_PATH
## Results
Some generation and completion results are as follows.
-
- 
-
Multimodal completion on a ShapeNet chair.
-
- 
-
Multimodal completion on PartNet.
-
- 
-
Multimodal completion on two Redwood 3DScan chairs.
-
- 
-
## Reference
-```
+```bibtex
@inproceedings{Zhou_2021_ICCV,
- author = {Zhou, Linqi and Du, Yilun and Wu, Jiajun},
- title = {3D Shape Generation and Completion Through Point-Voxel Diffusion},
- booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
- month = {October},
- year = {2021},
- pages = {5826-5835}
+ author = {Zhou, Linqi and Du, Yilun and Wu, Jiajun},
+ title = {3D Shape Generation and Completion Through Point-Voxel Diffusion},
+ booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
+ month = {October},
+ year = {2021},
+ pages = {5826-5835}
}
```
diff --git a/environment.yml b/environment.yml
new file mode 100644
index 0000000..92bad73
--- /dev/null
+++ b/environment.yml
@@ -0,0 +1,21 @@
+name: PVD
+
+channels:
+ - nodefaults
+ - drti
+ - pytorch
+ - nvidia
+ - conda-forge
+ - pyg
+ - open3d-admin
+
+dependencies:
+ - python
+ - pytorch
+ - torchvision
+ - cudatoolkit
+ - matplotlib
+ - open3d
+ - tqdm
+ - trimesh
+ - scipy
diff --git a/requirement_voxel.txt b/requirement_voxel.txt
deleted file mode 100644
index 9814675..0000000
--- a/requirement_voxel.txt
+++ /dev/null
@@ -1,26 +0,0 @@
-conda:
-python==3.6
-torch==1.4.0
-torchvision==0.5.0
-cudatoolkit==10.1
-kaolin==0.1.0
-pytorch3d==0.2.5
-lutorpy=1.3.7
-xmltodict=0.12.0
-numba=0.51.2
-pycuda=2019.1.2
-matplotlib
-
-pip:
-torch-scatter==2.0.4
-torch-sparse==0.6.1
-torch-cluster==1.5.4
-torch-spline-conv==1.2.0
-descartes==1.1.0
-fire==0.3.1
-jupyter==1.0.0
-opencv_python==4.3.0
-Shapely==1.7.0
-Pillow==6.2.1
-torch_geometric==1.6.0
-open3d