OpenTraj: Assessing Prediction Complexity in Human Trajectories Datasets

Javad Amirian, Bingqing Zhang, Francisco Valente Castro, Juan Jose Baldelomar, Jean-Bernard Hayet, Julien Pettre; Proceedings of the Asian Conference on Computer Vision (ACCV), 2020

Abstract

Human Trajectory Prediction (HTP) having gained much momentum in the last years, this paper addresses the question of evaluating how complex is a given dataset with respect to the prediction problem. For assessing a dataset complexity, we define a series of indicators around three concepts: Trajectory predictability; Trajectory regularity; Context complexity. We compare the most common datasets used in HTP at the light of these indicators and discuss what this may imply on benchmarking of HTP algorithms.

Related Material

[pdf] [arXiv] [code]
[bibtex]
@InProceedings{Amirian_2020_ACCV, author = {Amirian, Javad and Zhang, Bingqing and Castro, Francisco Valente and Baldelomar, Juan Jose and Hayet, Jean-Bernard and Pettre, Julien}, title = {OpenTraj: Assessing Prediction Complexity in Human Trajectories Datasets}, booktitle = {Proceedings of the Asian Conference on Computer Vision (ACCV)}, month = {November}, year = {2020} }