Step 1.
Prepare ScanNet 2D data. The processed 2D data can be downloaded from HERE. Run cat 2D.tar.* > 2D.tar to merge the files. Then skip to Step 2.
Or you can process ScanNet 2D data yourself by following the steps below.
First acquire download-scannet.py from HERE. You should fill out an agreement to the ScanNet Terms of Use.
Then download 2D data, run:
python download-scannet.py -o <ScanNet root> --type .sens
python download-scannet.py -o <ScanNet root> --type _2d-label.zip
python download-scannet.py -o <ScanNet root> --type _2d-instance.zip
Extract _2d-label.zip and _2d-instance.zip into 2D_info folder, whose structure follows:
2D_info
├── scenexxxx_xx.sens
├── scenexxxx_xx_2d-label/label/xxxxxx.png
└── scenexxxx_xx_2d-instance/instance/xxxxxx.png
Link the 2D_info folder to this level of directory.
Process 2D data by:
python prepare_2d_data.py --scannet_path ./2D_info --output_path ./2D --label_map_file ./meta_data/scannetv2-labels.combined.tsv --scene_index_file ./meta_data/scannet_train.txt
Step 2. Prepare ScanNet 3D data. Follow votenet to download and process the 3D data.
Link or move the scans folder to this level of directory.
Generate online data by:
python generate_online_data.py
Step 3. Generate .pkl files by:
python tools/create_data.py scannet --root-path ./data/scannet-mv --out-dir ./data/scannet-mv --extra-tag scannet_mv
Final folder structure:
scannet-mv
├── meta_data/
├── batch_load_scannet_data.py
├── load_scannet_data.py
├── scannet_utils.py
├── README.md
├── scans/
├── 2D_info/
├── 2D
│ ├── scenexxxx_xx
│ │ ├── color
│ │ │ ├── xxxxxx.jpg
│ │ ├── depth
│ │ │ ├── xxxxxx.png
│ │ ├── label
│ │ │ ├── xxxxxx.png
│ │ ├── instance
│ │ │ ├── xxxxxx.png
│ │ ├── pose
│ │ ├── xxxxxx.txt
├── generate_online_data.py
├── instance_mask
│ ├── scenexxxx_xx
│ │ ├── xxxxxx.npy
├── semantic_mask
│ ├── scenexxxx_xx
│ │ ├── xxxxxx.npy
├── point
│ ├── scenexxxx_xx
│ │ ├── xxxxxx.npy
├── modal_box
│ ├── scenexxxx_xx
│ │ ├── xxxxxx.npy
├── amodal_box_mask
│ ├── scenexxxx_xx
│ │ ├── xxxxxx.npy
├── scene_amodal_box
│ ├── scenexxxx_xx.npy
├── scannet_mv_infos_train.pkl
└── scannet_mv_infos_val.pkl