Published December 21, 2021 | Version v1
Dataset Open

Generated RGBD Tracking Benchmarks (COCO2017)

Authors/Creators

  • 1. Tampere University

Description

We convert the existing RGB tracking benchmakrs, eg. COCO, to RGB+pseudo-Depth tracking benchmarks using DenseDepth or HighResDepth monocular estimation methods.

Please visit the COCO page for the RGB images and the groundtruths.

Files

train2017.zip

Files (26.7 GB)

Name Size
md5:c25ee6511dc240cbe512ee5c42efce24
25.7 GB Preview Download
md5:391dc103fd7a55761cd7c7eff25bb6e8
1.1 GB Preview Download