Pedestrian Intention Estimation (PIE) Dataset

PIE is a new dataset for studying pedestrian behaviour in traffic. PIE contains over 6 hours of footage recorded in typical traffic scenes with on-board camera. It also provides accurate vehicle information from OBD sensor (vehicle speed, heading direction and GPS coordinates) synchronized with video footage. Rich spatial and behavioural annotations are available for pedestrians and vehicles that potentially interact with the ego-vehicle as well as for the relevant elements of infrastructure (traffic lights, signs and zebra crossings). There are over 300K labeled video frames with 1842 pedestrian samples making this the largest publicly available dataset for studying pedestrian behaviour in traffic.

Data and Resources

Additional Info

Field Value
Source http://data.nvision2.eecs.yorku.ca/PIE_dataset/
Author Amir Rasouli (aras@eecs.yorku.ca) and Iuliia Kotseruba (yulia_k@eecs.yorku.ca).
Maintainer Amir Rasouli (aras@eecs.yorku.ca) and Iuliia Kotseruba (yulia_k@eecs.yorku.ca).
Number of Instances 3
Package Description Total number of frames 909,480 Total number of annotated frames 293,437 Number of pedestrians with behavior annotations 1,842 Number of pedestrian bounding boxes 738,970 Number of traffic object bounding boxes 2,353,983 Average length of pedestrian track 401 frames Pedestrian counts Intend to cross and cross 519 Intend to cross and don't cross 894 Do not intend to cross 429
Dataset has missing values False