MTR-VP: Towards End-to-End Trajectory Planning through Context-Driven Image Encoding and Multiple Trajectory Prediction

TL;DR

This paper introduces MTR-VP, an end-to-end trajectory planning method for autonomous driving that uses vision-based context embeddings and multiple trajectory predictions, improving scene understanding and planning accuracy.

Contribution

The paper proposes a novel approach combining vision-based context encoding with multimodal trajectory prediction, replacing map features with learned visual representations in an end-to-end framework.

Findings

Predicting multiple future trajectories improves planning performance.

Transformer-based visual and kinetic feature integration has limitations.

Using learned visual scene context enhances trajectory prediction accuracy.

Abstract

We present a method for trajectory planning for autonomous driving, learning image-based context embeddings that align with motion prediction frameworks and planning-based intention input. Within our method, a ViT encoder takes raw images and past kinematic state as input and is trained to produce context embeddings, drawing inspiration from those generated by the recent MTR (Motion Transformer) encoder, effectively substituting map-based features with learned visual representations. MTR provides a strong foundation for multimodal trajectory prediction by localizing agent intent and refining motion iteratively via motion query pairs; we name our approach MTR-VP (Motion Transformer for Vision-based Planning), and instead of the learnable intention queries used in the MTR decoder, we use cross attention on the intent and the context embeddings, which reflect a combination of information encoded from the driving scene and past vehicle states. We evaluate our methods on the Waymo End-to-End Driving Dataset, which requires predicting the agent's future 5-second trajectory in bird's-eye-view coordinates using prior camera images, agent pose history, and routing goals. We analyze our architecture using ablation studies, removing input images and multiple trajectory output. Our results suggest that transformer-based methods that are used to combine the visual features along with the kinetic features such as the past trajectory features are not effective at combining both modes to produce useful scene context embeddings, even when intention embeddings are augmented with foundation-model representations of scene context from CLIP and DINOv2, but that predicting a distribution over multiple futures instead of a single future trajectory boosts planning performance.