Pedestrian Trajectory Prediction with Missing Data: Datasets, Imputation, and Benchmarking

TL;DR

This work comprehensively examine several imputation methods to reconstruct the missing coordinates and benchmark them for imputing pedestrian trajectories, providing a thorough analysis of recent trajectory prediction methods and evaluating the performance of these models on the imputed trajectories.

Abstract

Pedestrian trajectory prediction is crucial for several applications such as robotics and self-driving vehicles. Significant progress has been made in the past decade thanks to the availability of pedestrian trajectory datasets, which enable trajectory prediction methods to learn from pedestrians' past movements and predict future trajectories. However, these datasets and methods typically assume that the observed trajectory sequence is complete, ignoring real-world issues such as sensor failure, occlusion, and limited fields of view that can result in missing values in observed trajectories. To address this challenge, we present TrajImpute, a pedestrian trajectory prediction dataset that simulates missing coordinates in the observed trajectory, enhancing real-world applicability. TrajImpute maintains a uniform distribution of missing data within the observed trajectories. In this work, we comprehensively examine several imputation methods to reconstruct the missing coordinates and benchmark them for imputing pedestrian trajectories. Furthermore, we provide a thorough analysis of recent trajectory prediction methods and evaluate the performance of these models on the imputed trajectories. Our experimental evaluation of the imputation and trajectory prediction methods offers several valuable insights. Our dataset provides a foundational resource for future research on imputation-aware pedestrian trajectory prediction, potentially accelerating the deployment of these methods in real-world applications. Publicly accessible links to the datasets and code files are available at https://github.com/Pranav-chib/TrajImpute.

Figures (4)

• Figure 1: Number of trajectory samples in training, testing, and validation splits of the TrajImpute dataset.
• Figure 2: Illustration of the total missing coordinates in the easy and hard protocols for the ETH-M, HOTEL-M, UNIV-M, ZARA1-M, and ZARA2-M subsets of TrajImpute. 'M' refers to missing, indicating that the subset contains missing observed coordinates. The hard protocol creates more missing values compared to the easy protocol.
• Figure 3: Illustration of an example (right) showing the missing pattern in the trajectory sequence of the ETH-M subset under the easy protocol. TrajImpute ensures that coordinates are equally likely to be dropped from the trajectory sequence, ensuring that any time frame can result in missing observations in the past trajectory. Furthermore, the structure of the TrajImpute dataset contains a dictionary structure with eight keys (left). Here, N is the number of trajectory sequences.
• Figure 4: Illustrations of predictions under different observation conditions: clean, missing, soft imputed, and hard imputed observations.