matteorr / rel_3d_pose

A method for training a 3d pose estimation algorithm from relative depth annotations. Implemented in Pytorch.
MIT License
65 stars 7 forks source link

Mturk Data #9

Closed zhLawliet closed 2 years ago

zhLawliet commented 2 years ago

@matteorr where can i find the data?

Mturk Data:

We list the links to download the crowd collected relative depth annotations and the merged version which we release as the Relative Depth LSP Dataset. Download and put them in this directory to replicate the experiments from our paper.

Relative Depth LSP Dataset

Relative Depth LSP Raw Mturk Annotations

Relative Depth LSP Author Annotations

Relative Depth Human3.6m Mturk User Study

Relative Depth Human3.6m Ground Truth

elijahcole commented 2 years ago

Sorry for the trouble, we are transitioning our datasets to new hosting.

At the moment this data can be found here: https://data.caltech.edu/records/20096

zhLawliet commented 2 years ago

thanks