What type of study is this?

September 10, 2025SensorsOpen Access

Diff-Pre: A Diffusion Framework for Trajectory Prediction

View Full Paper

Ask AI

Bookmark

Authors

This framework enhances trajectory prediction accuracy in vehicles using a diffusion model and attention mechanisms.

The proposed model achieved significant improvements in trajectory prediction accuracy under complex conditions.
Average displacement error (ADE) reached 0.199 m, and final displacement error (FDE) was 0.437 m over 1 to 5 seconds.
This model employs long short-term memory networks and multi-head attention to dynamically capture vehicle interactions.
It demonstrates superior robustness and predictive accuracy in high-interaction scenarios like lane changes.

Liu et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark