Spatio-Temporal Retrieval-based Priors for Adaptive Computational Teaching in Driving

2026-06-23

AI Summary

This work addresses the challenge that existing learning-based driving instruction systems struggle to model learners’ long-term learning dynamics and the cumulative effects of teacher–student interactions, thereby limiting adaptive teaching efficacy. To overcome this, the authors propose an imitation learning–based adaptive teaching model that innovatively integrates spatiotemporal nearest-neighbor retrieval with cross-attention priors to construct a temporal reasoning module tailored for repetitive instructional tasks. This module enables effective inference under low-data regimes by leveraging semantically similar segments from historical interactions. Experimental results demonstrate that the proposed method significantly outperforms both non-adaptive and adaptive baselines on a semi-synthetic longitudinal interaction dataset and a real-world racing simulation teaching dataset, validating its effectiveness in mitigating interaction data scarcity and enhancing instructional adaptability.

Abstract

Learning-based automated coaching systems for complex motor tasks such as high-performance driving remain limited in the ability to be adaptive by their reliance only on local, context-dependent reasoning, failing to account for the long-term temporal nature of student learning and the cumulative impact of repeated teacher-student interactions. In this paper, we propose an imitation learning based computational model for adaptive teaching with a dedicated temporal reasoning module that can reason over the interaction history under low-data regimes. To compensate for limited amounts of interactive training data, and based on the repetitive nature of the teaching process, the model relies on a nearest neighbor retrieval and cross attention prior, reasoning only on a narrowed-down set of semantically similar past interactions with an encoder-decoder based concurrent teaching model. We validate our approach with (i) a novel semi-synthetic closed-loop longitudinal student-teacher interaction dataset based on Waymo Open Motion Dataset and (ii) a small-scale real-world naturalistic simulator race coaching dataset. Our results reveal the consistent advantage of our adaptive teaching model with the nearest neighbor retrieval and cross-attention prior over a non-adaptive baseline as well as a suite of adaptive models that differ in their choice of priors and temporal fusion mechanisms.

Problem

• adaptive teaching
• spatio-temporal reasoning
• imitation learning
• computational coaching
• long-term interaction

Innovation

• adaptive teaching
• spatio-temporal retrieval
• cross-attention prior
• imitation learning
• low-data regime