Abstract:The objective of pedestrian trajectory prediction is to estimate the future paths of pedestrians by leveraging historical observations, which plays a vital role in ensuring the safety of self-driving vehicles and navigation robots. Previous works usually rely on a sufficient amount of observation time to accurately predict future trajectories. However, there are many real-world situations where the model lacks sufficient time to observe, such as when pedestrians abruptly emerge from blind spots, resulting in inaccurate predictions and even safety risks. Therefore, it is necessary to perform trajectory prediction based on instantaneous observations, which has rarely been studied before. In this paper, we propose a Bi-directional Consistent Diffusion framework tailored for instantaneous trajectory prediction, named BCDiff. At its heart, we develop two coupled diffusion models by designing a mutual guidance mechanism which can bidirectionally and consistently generate unobserved historical trajectories and future trajectories step-by-step, to utilize the complementary information between them. Specifically, at each step, the predicted unobserved historical trajectories and limited observed trajectories guide one diffusion model to generate future trajectories, while the predicted future trajectories and observed trajectories guide the other diffusion model to predict unobserved historical trajectories. Given the presence of relatively high noise in the generated trajectories during the initial steps, we introduce a gating mechanism to learn the weights between the predicted trajectories and the limited observed trajectories for automatically balancing their contributions. By means of this iterative and mutually guided generation process, both the future and unobserved historical trajectories undergo continuous refinement, ultimately leading to accurate predictions. Essentially, BCDiff is an encoder-free framework that can be compatible with existing trajectory prediction models in principle. Experiments show that our proposed BCDiff significantly improves the accuracy of instantaneous trajectory prediction on the ETH/UCY and Stanford Drone datasets, compared to related approaches.

TrajDiffuse: A Conditional Diffusion Model for Environment-Aware Trajectory Prediction

Enhanced Multimodal Trajectory Prediction for Autonomous Vehicles Using Advanced Diffusion Model Techniques

DICE: Diverse Diffusion Model with Scoring for Trajectory Prediction

Diffusion-Based Environment-Aware Trajectory Prediction

DiffTraj: Generating GPS Trajectory with Diffusion Probabilistic Model

DiffuTraj: A Stochastic Vessel Trajectory Prediction Approach via Guided Diffusion Process

Intention-aware Denoising Diffusion Model for Trajectory Prediction

Controllable Diverse Sampling for Diffusion Based Motion Behavior Forecasting

Trajectory Grid Diffusion for Multimodal Trajectory Prediction in Autonomous Vehicles

Diff-RNTraj: A Structure-aware Diffusion Model for Road Network-constrained Trajectory Generation

BCDiff: Bidirectional Consistent Diffusion for Instantaneous Trajectory Prediction.

Motion Latent Diffusion for Stochastic Trajectory Prediction.

Constraint-Aware Diffusion Models for Trajectory Optimization

Predicting trajectory destinations based on diffusion model integrating spatiotemporal features and urban contexts

A multi-modal vehicle trajectory prediction framework via conditional diffusion model: A coarse-to-fine approach

Characterized Diffusion Networks for Enhanced Autonomous Driving Trajectory Prediction

Stochastic Trajectory Prediction Via Motion Indeterminacy Diffusion

EquiDiff: A Conditional Equivariant Diffusion Model For Trajectory Prediction

ControlTraj: Controllable Trajectory Generation with Topology-Constrained Diffusion Model

DiffTAD: Denoising diffusion probabilistic models for vehicle trajectory anomaly detection