Exploring the dynamics of head pointing

Target pointing is a core component of human-computer interaction, often coupled with a selection mechanism. Head-based pointing systems enable hands-free interaction, making them particularly useful for users who cannot rely on manual input. These systems generally offer high accuracy but at the cost of moderate interaction speed. Efforts to increase speed can lead to overshooting, necessitating a balance between speed and accuracy. Despite the growing interest in head-pointing systems, existing research has not addressed overshooting or undershooting, particularly in terms of their measurement and mitigation.

In this study, we conducted a user experiment to evaluate head-pointing performance in a two-dimensional environment, comparing three methods incorporating static and dynamic Control Display gains. We introduced definitions to identify and measure over- and undershooting in such contexts. Our results did not support the hypothesis that dynamic gain reduces over- or undershooting or decreases their likelihood. Instead, the low static gain method enabled the fastest and most accurate selections. The recorded data is available in the Stuttgart University data repository.