BATON is a 127-driver, 136.6-hour naturalistic dataset of both manual-to-automation handovers and automation-to-manual takeovers, and its baselines show multimodal fusion beats video-only for predicting them.
Title resolution pending
2 Pith papers cite this work, alongside 658 external citations. Polarity classification is still indexing.
2
Pith papers citing it
658
external citations · OpenAlex
years
2026 2representative citing papers
Response maps from game theory are used to model and differentiate human driver acceleration behaviors in response to yielding, non-yielding, and responsive AV actions based on simulator experiments.
citing papers explorer
-
BATON: A Multimodal Benchmark for Bidirectional Automation Transition Observation in Naturalistic Driving
BATON is a 127-driver, 136.6-hour naturalistic dataset of both manual-to-automation handovers and automation-to-manual takeovers, and its baselines show multimodal fusion beats video-only for predicting them.
-
Characterizing Driver Interactions with Autonomous Vehicles via Response Maps
Response maps from game theory are used to model and differentiate human driver acceleration behaviors in response to yielding, non-yielding, and responsive AV actions based on simulator experiments.