PersonaDrive: Controllable Trajectory Prediction with Multi-Dimensional Driving Personas

2026-08-15

AI Summary

This study addresses the lack of multidimensional controllability in autonomous driving trajectory prediction and the inability of existing benchmarks to distinguish driving styles under similar urgency levels. To overcome these limitations, we propose the PersonaDrive framework and introduce the PCT dataset. Our approach employs dual-axis persona decomposition and language modeling to capture multidimensional driving characteristics, integrating PCAT hierarchical anchor transformation, BEV-level multimodal fusion, and a specialized anti-collapse loss function to enable personalized trajectory generation. Experimental results demonstrate that PersonaDrive significantly outperforms baseline methods across diverse scenarios, effectively enhancing the accuracy, diversity, and controllability of trajectory prediction.

Abstract

Although recent trajectory prediction and end-to-end autonomous driving methods improve robustness in urban environments, they still lack meaningful controllability. Existing benchmarks either provide no persona-conditioned annotations or support only a single urgency spectrum (i.e., emergency, normal, relaxed), which cannot distinguish personas that share the same urgency level but require different driving dynamics. To address this, we propose (i) the Persona-Conditioned Trajectory (PCT) dataset, which decomposes driving personas along two axes, Temporal Urgency and Ride Comfort, and combines three levels of each to form a grid of nine personas, each paired with natural-language descriptions and trajectories, and (ii) PersonaDrive, a framework that can learn driving personas from language and can generate persona-specific trajectories. PersonaDrive incorporates Persona-Conditioned Anchor Transform (PCAT), which hierarchically reshapes anchors along both axes, and Persona-Conditioned Multi-Modal Fusion (PCMF) for BEV-level persona fusion. Training is supervised by a Hierarchical Guide Loss enforcing axis-aligned physical orderings and an Axis-Decomposed Diversity Loss preventing diagonal mode collapse. Experimental results show that PersonaDrive consistently improves over the compared baselines across multi-dimensional scenarios. The code and PCT dataset are available at https://github.com/VisualAIKHU/PersonaDrive

Problem

• Trajectory Prediction
• Controllability
• Driving Personas
• Autonomous Driving
• Benchmark

Innovation

• Controllable Trajectory Prediction
• Multi-Dimensional Driving Personas
• Persona-Conditioned Anchor Transform
• Hierarchical Guide Loss
• Axis-Decomposed Diversity Loss