Domain Knowledge Driven Pseudo Labels for Interpretable Goal-Conditioned Interactive Trajectory Prediction

TL;DR

This paper introduces a goal-conditioned trajectory prediction model using a CVAE with pseudo labels to encode interaction modes, improving interpretability and addressing issues like posterior collapse in autonomous driving scenarios.

Contribution

The paper proposes a novel pseudo label approach to induce an interpretable latent space in CVAE models for joint trajectory prediction, incorporating domain knowledge effectively.

Findings

Addresses posterior collapse with pseudo labels

Validates approach on Waymo dataset with positive results

Enhances interpretability of interaction modes in trajectory prediction

Abstract

Motion forecasting in highly interactive scenarios is a challenging problem in autonomous driving. In such scenarios, we need to accurately predict the joint behavior of interacting agents to ensure the safe and efficient navigation of autonomous vehicles. Recently, goal-conditioned methods have gained increasing attention due to their advantage in performance and their ability to capture the multimodality in trajectory distribution. In this work, we study the joint trajectory prediction problem with the goal-conditioned framework. In particular, we introduce a conditional-variational-autoencoder-based (CVAE) model to explicitly encode different interaction modes into the latent space. However, we discover that the vanilla model suffers from posterior collapse and cannot induce an informative latent space as desired. To address these issues, we propose a novel approach to avoid KL vanishing and induce an interpretable interactive latent space with pseudo labels. The proposed pseudo labels allow us to incorporate domain knowledge on interaction in a flexible manner. We motivate the proposed method using an illustrative toy example. In addition, we validate our framework on the Waymo Open Motion Dataset with both quantitative and qualitative evaluations.