File size: 1,807 Bytes
41c75ec | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 | # Prepare SceneFlow dataset
Scene Flow Datasets is available at the Scene Flow [official website](https://lmb.informatik.uni-freiburg.de/resources/datasets/SceneFlowDatasets.en.html).
Please download the following parts of FlyingThings3D, Driving, Monkaa and place it in the `datasets/sceneflow` directory.
- RGB images (finalpass)
- RGB images (cleanpass)
- Disparity
And for FlyingThings3D subset, you need to download the following parts:
- RGB images (cleanpass)
- Disparity
- Disparity Occlusions
The directory structure should be as follows:
```text
data
| SceneFlow
| ├── Driving
| | ├── frames_finalpass
| | ├── frames_cleanpass
| | ├── disparity
...
| ├── Monkaa
| | ├── frames_finalpass
| | ├── frames_cleanpass
| | ├── disparity
...
| ├── FlyingThings3D
| | ├── frames_finalpass
| | ├── frames_cleanpass
| | ├── disparity
...
| ├── FlyingThings3D_subset
| | ├── train
| | | ├── frames_cleanpass
| | | ├── disparity
| | | ├── disparity_occlusions
| | ├── test
| | | ├── frames_cleanpass
| | | ├── disparity
| | | ├── disparity_occlusions
```
_Optionally you can write your own txt file and use all the parts of the dataset._
### Reference
```bibtex
@inproceedings{mayer2016large,
title={A large dataset to train convolutional networks for disparity, optical flow, and scene flow estimation},
author={Mayer, Nikolaus and Ilg, Eddy and Hausser, Philip and Fischer, Philipp and Cremers, Daniel and Dosovitskiy, Alexey and Brox, Thomas},
booktitle={Proceedings of the IEEE conference on computer vision and pattern recognition},
pages={4040--4048},
year={2016}
}
``` |