five

RobustSpring: Benchmarking Robustness to Image Corruptions for Optical Flow, Scene Flow and Stereo

收藏
DataCite Commons2025-05-20 更新2025-05-18 收录
下载链接:
https://darus.uni-stuttgart.de/citation?persistentId=doi:10.18419/DARUS-5047
下载链接
链接失效反馈
官方服务:
资源简介:
The RobustSpring dataset contains the image corruption data files for scene flow, optical flow and stereo estimation with the <a href="https://doi.org/10.18419/DARUS-3376">Spring dataset</a>. Note that this repository contains only the Spring test data files. For easier handling, we organized them into sub-directories by image corruption type: <br> <ul> <li><code>brightness.zip </code>: brightness image corruption</li> <li><code>contrast.zip </code>: contrast image corruption</li> <li><code>defocus_blur.zip </code>: defocus_blur image corruption</li> <li><code>elastic_transform.zip </code>: elastic_transform image corruption</li> <li><code>fog.zip </code>: fog image corruption</li> <li><code>frost.zip </code>: frost image corruption</li> <li><code>gaussian_blur.zip </code>: gaussian_blur image corruption</li> <li><code>gaussian_noise.zip </code>: gaussian_noise image corruption</li> <li><code>glass_blur.zip </code>: glass_blur image corruption</li> <li><code>impulse_noise.zip </code>: impulse_noise image corruption</li> <li><code>jpeg_compression.zip </code>: jpeg_compression image corruption</li> <li><code>motion_blur.zip </code>: motion_blur image corruption</li> <li><code>pixelate.zip </code>: pixelate image corruption</li> <li><code>rain.zip </code>: rain image corruption</li> <li><code>saturate.zip </code>: saturate image corruption</li> <li><code>shot_noise.zip </code>: shot_noise image corruption</li> <li><code>snow.zip </code>: snow image corruption</li> <li><code>spatter.zip </code>: spatter image corruption</li> <li><code>speckle_noise.zip </code>: speckle_noise image corruption</li> <li><code>zoom_blur.zip </code>: zoom_blur image corruption</li> </ul> <br> Each image corruption folder is internally organized as follows: <br> <ul> <li><code>test </code>: Indicates that this is the test proportion of the Spring dataset</li> <ul> <li><code>0003</code>: Scene subfolder for scene 0003, 131 frames</li> <ul> <li><code>frame_left</code>: Left frames</li> <ul> <li><code>frame_left_0001.png</code>: Frame 0001</li> <li>...</li> <li><code>frame_left_0131.png</code>: Frame 0131</li> </ul> <li><code>frame_right</code>: Right frames</li> <ul> <li><code>frame_right_0001.png</code>: Frame 0001</li> <li>...</li> <li><code>frame_right_0131.png</code>: Frame 0131</li> </ul> </ul> <li><code>0019</code>: Scene 0019, same internal structure as above, 111 frames.</li> <li><code>0028</code>: Scene 0028, same internal structure as above, 39 frames.</li> <li><code>0028</code>: Scene 0029, same internal structure as above, 135 frames.</li> <li><code>0031</code>: Scene 0031, same internal structure as above, 73 frames.</li> <li><code>0034</code>: Scene 0034, same internal structure as above, 47 frames.</li> <li><code>0035</code>: Scene 0035, same internal structure as above, 120 frames.</li> <li><code>0040</code>: Scene 0040, same internal structure as above, 111 frames.</li> <li><code>0042</code>: Scene 0042, same internal structure as above, 116 frames.</li> <li><code>0046</code>: Scene 0046, same internal structure as above, 117 frames.</li> </ul> </ul> <br> File Formats: <br> <ul> <li>All images are given as .png files</li> </ul> <br> For the project website see <a href="https://spring-benchmark.org">spring-benchmark.org</a>.
提供机构:
DaRUS
创建时间:
2025-05-13
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作