People Approaching Robots Database (PAR-D)

PAR-D is a database containing two datasets about people approaching robots in public spaces:

• The Shutter Interaction Dataset
• Augmented labels for the ATC Trajectory Dataset - please additionally cite the original ATC Trajectory Dataset paper if using the augmented labels with the original data

You can learn more about the Shutter Interaction Dataset here and in the Shutter Interaction Dataset README contained in Shutter_Interaction_Dataset/README_shutter.md.

If you would like to reproduce the results from the paper using PAR-D, you can find the code for the paper here.

PAR-D was introduced in the paper Predicting Human Intent to Interact with a Public Robot: The People Approaching Robots Database (PAR-D). Please cite this paper if you use elements of PAR-D:

Thompson, Sydney, Alexander Lew, Yifan Li, Elizabeth Stanish, Alex Huang, Rohan Phanse, and Marynel Vázquez. "Predicting Human Intent to Interact with a Public Robot: The People Approaching Robots Database (PAR-D)." In Proceedings of the 26th International Conference on Multimodal Interaction, pp. 536-545. 2024.

Contributing to PAR-D

If you have data that you think would compliment PAR-D well, please contact Sydney Thompson at <firstname>.<lastname>@yale.edu so we can chat!