Uncertainty-Aware Trajectory Prediction via Rule-Regularized Heteroscedastic Deep Classification
This addresses robustness and calibration issues in trajectory prediction for autonomous driving, particularly in complex scenarios like intersections, but it is incremental as it builds on existing methods with novel hybrid techniques.
The paper tackles the challenge of out-of-distribution generalization in deep learning-based trajectory prediction by proposing SHIFT, a framework that combines uncertainty modeling with rule-based priors, achieving substantial gains in uncertainty calibration and displacement metrics on the nuScenes dataset.
Deep learning-based trajectory prediction models have demonstrated promising capabilities in capturing complex interactions. However, their out-of-distribution generalization remains a significant challenge, particularly due to unbalanced data and a lack of enough data and diversity to ensure robustness and calibration. To address this, we propose SHIFT (Spectral Heteroscedastic Informed Forecasting for Trajectories), a novel framework that uniquely combines well-calibrated uncertainty modeling with informative priors derived through automated rule extraction. SHIFT reformulates trajectory prediction as a classification task and employs heteroscedastic spectral-normalized Gaussian processes to effectively disentangle epistemic and aleatoric uncertainties. We learn informative priors from training labels, which are automatically generated from natural language driving rules, such as stop rules and drivability constraints, using a retrieval-augmented generation framework powered by a large language model. Extensive evaluations over the nuScenes dataset, including challenging low-data and cross-location scenarios, demonstrate that SHIFT outperforms state-of-the-art methods, achieving substantial gains in uncertainty calibration and displacement metrics. In particular, our model excels in complex scenarios, such as intersections, where uncertainty is inherently higher. Project page: https://kumarmanas.github.io/SHIFT/.