The study of human navigation has long been of interest to various research communities such as computer graphics, computer vision, cognitive science, and robotics. Advancements in these areas have seen widespread practical application in pandemic response, architectural design, urban planning, transportation engineering, crowd management, socially compliant robot navigation, and entertainment. Accordingly, the influence of human navigation research has reached countless 8 individuals and will continue to do so in the foreseeable future.

Datasets for Human Trajectory Prediction

Existing human trajectory datasets have limitations in the sense of embodying interactions. They either do not contain Agent-to-Environment (A2E) interactions, or exhibit limited Agent-to-Agent (A2A) interactions at small scale in simple environments. We speculate that many self-centered pedestrians are prone to avoid or mitigate, consciously or unconsciously, the influence of the environments and other pedestrians during their navigation.

In this work, we are proposing datasets that augment A2E and A2A interactions, which may bring benefits for enhancing learning models by encoding more complex dynamics in trajectories.

We propose a comprehensive trajectory prediction dataset A-to-X that consists of a representative set of trajectories, which will enable better generalization under realistic circumstances that are either complex or unsafe and out-of-distribution (OOD) with respect to current datasets.