Exploring the Impact of Hand Pose and Shadow on Hand-washing Action Recognition
Shengtai Ju, Amy R. Reibman
TL;DR
The paper addresses distribution shift in camera-based handwashing action recognition caused by hand pose and outdoor shadow. It leverages synthetic data with controlled pose and shadow variations to quantify breakdown points and evaluate robustness using a lightweight image classifier. Key findings show that pose can trigger sharp performance drops beyond certain angles, shadow intensity drives larger accuracy losses than shadow size, and adding training poses in the range $50^{\circ}$–$60^{\circ}$ most effectively mitigates breakdown points. The work contributes a quantitative characterization of breakdown points, a shadow-robustness analysis, and a simple pose-augmentation strategy, with implications for outdoor handwashing recognition systems.
Abstract
In the real world, camera-based application systems can face many challenges, including environmental factors and distribution shift. In this paper, we investigate how pose and shadow impact a classifier's performance, using the specific application of handwashing action recognition. To accomplish this, we generate synthetic data with desired variations to introduce controlled distribution shift. Using our synthetic dataset, we define a classifier's breakdown points to be where the system's performance starts to degrade sharply, and we show these are heavily impacted by pose and shadow conditions. In particular, heavier and larger shadows create earlier breakdown points. Also, it is intriguing to observe model accuracy drop to almost zero with bigger changes in pose. Moreover, we propose a simple mitigation strategy for pose-induced breakdown points by utilizing additional training data from non-canonical poses. Results show that the optimal choices of additional training poses are those with moderate deviations from the canonical poses with 50-60 degrees of rotation.