Dense temporal annotation of procedural activity videos is vital for action understanding and embodied intelligence but remains labor-intensive due to reactive tools. Each correction is treated as an isolated edit, limiting reuse of information on annotator uncertainty and model reliability. We introduce IMPACT-Scribe, a correction-driven framework for dense labeling that uses each correction to improve future human-machine collaboration. IMPACT-Scribe combines uncertainty-aware boundary scribble supervision, local proposal modeling, cost-aware query planning, structured propagation, and correction-driven adaptation. Experiments and a human study show that this closed-loop design improves labeling quality per effort, enhances boundary accuracy, and fosters better human-machine interaction over time. The code will be made publicly available at https://github.com/BanzQians/IMPACT_AS.
@article{arxiv.2605.01668,
title = {IMPACT-Scribe: Interactive Temporal Action Segmentation with Boundary Scribbles and Query Planning},
author = {Qian Yin and Di Wen and Kunyu Peng and David Schneider and Zeyun Zhong and Alexander Jaus and Zdravko Marinov and Jiale Wei and Ruiping Liu and Junwei Zheng and Yufan Chen and Chen Zhang and Lei Qi and Rainer Stiefelhagen},
journal= {arXiv preprint arXiv:2605.01668},
year = {2026}
}
Comments
7 pages, 4 figures. Code is available at https://github.com/BanzQians/IMPACT_AS