Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Multi-Stage Goal-Driven Network for Pedestrian Trajectory Prediction

Xiuen Wu, Tao Wang, Yuanzheng Cai, Lingyu Liang, George Papageorgiou

TL;DR

Diverging from prior approaches relying on stepwise recursive prediction and the singular forecasting of a long-term goal, MGNet directs trajectory generation by forecasting intermediate stage goals, thereby reducing prediction errors.

Abstract

Pedestrian trajectory prediction plays a pivotal role in ensuring the safety and efficiency of various applications, including autonomous vehicles and traffic management systems. This paper proposes a novel method for pedestrian trajectory prediction, called multi-stage goal-driven network (MGNet). Diverging from prior approaches relying on stepwise recursive prediction and the singular forecasting of a long-term goal, MGNet directs trajectory generation by forecasting intermediate stage goals, thereby reducing prediction errors. The network comprises three main components: a conditional variational autoencoder (CVAE), an attention module, and a multi-stage goal evaluator. Trajectories are encoded using conditional variational autoencoders to acquire knowledge about the approximate distribution of pedestrians' future trajectories, and combined with an attention mechanism to capture the temporal dependency between trajectory sequences. The pivotal module is the multi-stage goal evaluator, which utilizes the encoded feature vectors to predict intermediate goals, effectively minimizing cumulative errors in the recursive inference process. The effectiveness of MGNet is demonstrated through comprehensive experiments on the JAAD and PIE datasets. Comparative evaluations against state-of-the-art algorithms reveal significant performance improvements achieved by our proposed method.

A Multi-Stage Goal-Driven Network for Pedestrian Trajectory Prediction

TL;DR

Diverging from prior approaches relying on stepwise recursive prediction and the singular forecasting of a long-term goal, MGNet directs trajectory generation by forecasting intermediate stage goals, thereby reducing prediction errors.

Abstract

Pedestrian trajectory prediction plays a pivotal role in ensuring the safety and efficiency of various applications, including autonomous vehicles and traffic management systems. This paper proposes a novel method for pedestrian trajectory prediction, called multi-stage goal-driven network (MGNet). Diverging from prior approaches relying on stepwise recursive prediction and the singular forecasting of a long-term goal, MGNet directs trajectory generation by forecasting intermediate stage goals, thereby reducing prediction errors. The network comprises three main components: a conditional variational autoencoder (CVAE), an attention module, and a multi-stage goal evaluator. Trajectories are encoded using conditional variational autoencoders to acquire knowledge about the approximate distribution of pedestrians' future trajectories, and combined with an attention mechanism to capture the temporal dependency between trajectory sequences. The pivotal module is the multi-stage goal evaluator, which utilizes the encoded feature vectors to predict intermediate goals, effectively minimizing cumulative errors in the recursive inference process. The effectiveness of MGNet is demonstrated through comprehensive experiments on the JAAD and PIE datasets. Comparative evaluations against state-of-the-art algorithms reveal significant performance improvements achieved by our proposed method.
Paper Structure (18 sections, 7 equations, 4 figures, 3 tables)

This paper contains 18 sections, 7 equations, 4 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Trajectory Prediction in Dynamic Video Scenes
  4. Attention-based Methods for Trajectory Prediction
  5. Goal-Driven Methods for Trajectory Prediction
  6. Proposed Method
  7. Problem Formulation
  8. Overview
  9. Temporal Attention
  10. Conditional Variational Autoencoder
  11. Multi-Stage Goal Estimator
  12. Loss Functions
  13. Experiments
  14. Datasets
  15. Implementation Details
  16. ...and 3 more sections

Figures (4)

  • Figure 1: Example of trajectory prediction results. The blue line represents the past trajectory, the orange line represents the actual future trajectory, and the green line represents the predicted trajectory.
  • Figure 2: An overview of our MGNet Architecture. arrows in orange, green, and blue denote connections during training, inference, and both training and inference, respectively.
  • Figure 3: Detailed structure of Multi-Stage Goal Evaluator. It adopts a double-layer structure, and uses the stage goal features predicted by the upper layer to guide the generation of lower-level stage goals, and outputs several stage goal features.
  • Figure 4: Qualitative results on trajectory prediction on the JAAD dataset. The ground truth is orange and the predictions is green. Best viewed in color.