The Datasets should be organized in the following way:
data/
├───data_data_CB/ # Cornell-Box Dataset
│ ├───*_renders_scene_*/ # folders for individual scenes
│ ├───...
│ ├───train.txt
│ ├───test.txt
│ └───val.txt
|
├───data_agresti/ # Datasets from Agresti et al.
│ ├───S1/
│ │ └───synthetic_dataset/
│ │ ├───test_set/
│ │ │ └───ground_truth/
│ │ └───training_set/
│ │ └───ground_truth/
│ ├───S2/
│ │ └───S2/
│ ├───S3/
│ │ └───S3/
│ │ └───ground_truth/
│ ├───S4/
│ │ └───real_dataset/
│ │ └───ground_truth/
│ └───S5/
│ └───S5/
│ └───ground_truth/
|
└───data_FLAT/ # FLAT Dataset
├───kinect/
│ ├───full/
│ ├───gt/
│ ├───list/
│ └───msk/
├───phasor/ # not used in the code
└───deeptof/ # not used in the code
Other paths/path_structures may be specified in the respective data_loader.py files.