About the dataset
#2
by
HongzheBi
- opened
I noticed that the current dataset contains 1.7k trajectories, while the paper mentions having 27k human trajectories. Is it possible to open source the entire dataset?
Thank you for your question. We have some external collaborators who helped collect (a large part of) the dataset. They have licensing requirement on their side of the data and we are working through the process.
This dataset will be a continuing effort. In addition, we are also working to open-source implementations to collect human data with VR devices.