Driving Stereo Datasets is available at the Driving Stereo official website.
Please download the dataset from Google Drive or BaiduCloud and unzip it to the data
directory.
To use our list file for training and testing, the directory structure should be as follows:
data
├── DrivingStereo
| ├── calib_test
| ├── calib_train
| ├── cloudy
| ├── foggy
| ├── rainy
| ├── sunny
| ├── test-depth-map
| ├── test-disparity-map
| ├── test-left-image
| ├── test-right-image
| ├── train-depth-map
| ├── train-disparity-map
| ├── train-left-image
| ├── train-right-image
Optionally you can write your own txt file and use all the parts of the dataset.
@inproceedings{yang2019drivingstereo,
title={DrivingStereo: A Large-Scale Dataset for Stereo Matching in Autonomous Driving Scenarios},
author={Yang, Guorun and Song, Xiao and Huang, Chaoqin and Deng, Zhidong and Shi, Jianping and Zhou, Bolei},
booktitle={IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
year={2019}
}