IMPACT-HOI: Supervisory Control for Onset-Anchored Partial HOI Event Construction

TL;DR

IMPACT-HOI is a mixed-initiative framework for annotating egocentric videos by constructing structured event graphs for HOI, improving annotation efficiency and accuracy with a risk-bounded protocol.

Contribution

It introduces a novel incremental, onset-anchored event construction method with a trust-calibrated controller and risk-bounded execution for high-quality supervision.

Findings

Reduced manual annotation actions by 13.5%.

Achieved a 46.67% event match rate.

Zero confirmed-field violations in the user study.

Abstract

We present IMPACT-HOI, a mixed-initiative framework for annotating egocentric procedural video by constructing structured event graphs for Human-Object Interactions (HOI), motivated by the need for high-quality structured supervision for learning robot manipulation from human demonstration. IMPACT-HOI frames this task as the incremental resolution of a partially specified, onset-anchored event state. A trust-calibrated controller selects among direct queries, human-confirmed suggestions, and conservative completions based on empirical annotator behavior and evidence quality. A risk-bounded execution protocol, utilizing atomic rollback, ensures that human-confirmed decisions are preserved against conflicting automated updates. A user study with 9 participants shows a 13.5% reduction in manual annotation actions, a 46.67% event match rate, and zero confirmed-field violations under the studied protocol. The code will be made publicly available at https://github.com/541741106/IMPACT_HOI.