The overall directory structure should be:
│ReCon/
├──cfgs/
├──data/
│ ├──ModelNet/
│ ├──ModelNetFewshot/
│ ├──ScanObjectNN/
│ ├──ShapeNet55-34/
│ ├──shapenetcore_partanno_segmentation_benchmark_v0_normal/
├──datasets/
├──.......
│ShapeNet55-34/
├──shapenet_pc/
│ ├── 02691156-1a04e3eab45ca15dd86060f189eb133.npy
│ ├── 02691156-1a6ad7a24bb89733f412783097373bdc.npy
│ ├── .......
├──shapenet_img/
│ ├── 02691156-1a04e3eab45ca15dd86060f189eb133.png
│ ├── 02691156-1a6ad7a24bb89733f412783097373bdc.png
│ ├── .......
├──ShapeNet-55/
│ ├── train.txt
│ └── test.txt
├──shapenet55.json
Download:
Please download the ShapeNet55 pointcloud data from Point-BERT repo, and download the ShapeNet55 image data from Our Google Drive. Specially, the image data is different from the pointcloud data in some samples, you need to update the meta-data "ShapeNet55-34/ShapeNet-55/train.txt & test.txt" from Our Google Drive.
│ModelNet/
├──modelnet40_normal_resampled/
│ ├── modelnet40_shape_names.txt
│ ├── modelnet40_train.txt
│ ├── modelnet40_test.txt
│ ├── modelnet40_train_8192pts_fps.dat
│ ├── modelnet40_test_8192pts_fps.dat
Download: You can download the processed data from Point-BERT repo, or download from the official website and process it by yourself.
│ModelNetFewshot/
├──5way10shot/
│ ├── 0.pkl
│ ├── ...
│ ├── 9.pkl
├──5way20shot/
│ ├── ...
├──10way10shot/
│ ├── ...
├──10way20shot/
│ ├── ...
Download: Please download the data from Point-BERT repo. We use the same data split as theirs.
│ScanObjectNN/
├──main_split/
│ ├── training_objectdataset_augmentedrot_scale75.h5
│ ├── test_objectdataset_augmentedrot_scale75.h5
│ ├── training_objectdataset.h5
│ ├── test_objectdataset.h5
├──main_split_nobg/
│ ├── training_objectdataset.h5
│ ├── test_objectdataset.h5
Download: Please download the data from the official website.
|shapenetcore_partanno_segmentation_benchmark_v0_normal/
├──02691156/
│ ├── 1a04e3eab45ca15dd86060f189eb133.txt
│ ├── .......
│── .......
│──train_test_split/
│──synsetoffset2category.txt
Download: Please download the data from here.