When to Act: Calibrated Confidence for Reliable Human Intention Prediction in Assistive Robotics
This addresses safety risks in assistive devices for users with disabilities by improving reliability in intention prediction, though it is incremental as it builds on existing calibration techniques.
The paper tackled the problem of unreliable confidence in predicting user intentions for assistive robotics, introducing a calibration method that reduces miscalibration by about an order of magnitude without affecting accuracy, enabling a safety-critical triggering framework.
Assistive devices must determine both what a user intends to do and how reliable that prediction is before providing support. We introduce a safety-critical triggering framework based on calibrated probabilities for multimodal next-action prediction in Activities of Daily Living. Raw model confidence often fails to reflect true correctness, posing a safety risk. Post-hoc calibration aligns predicted confidence with empirical reliability and reduces miscalibration by about an order of magnitude without affecting accuracy. The calibrated confidence drives a simple ACT/HOLD rule that acts only when reliability is high and withholds assistance otherwise. This turns the confidence threshold into a quantitative safety parameter for assisted actions and enables verifiable behavior in an assistive control loop.