Spmark: improving small-target text selection with smooth-pursuit gaze interaction in hybrid paper–digital environments

Precise text annotation on printed documents is essential for knowledge capture. While gaze-based interaction offers a hands-free approach, existing methods often suffer from low accuracy and ambiguous intent recognition in cluttered contexts. We propose SPMark, a hybrid gaze interaction method combining smooth pursuit and spatial proximity to support accurate text selection and annotation in hybrid paper-digital environments. SPMark introduces two mechanisms: an Interaction Area Activation Mechanism that dynamically restricts visible interaction objects based on gaze proximity, and a Hybrid Intent Locking Mechanism that fuses gaze distance and pursuit correlation to disambiguate intent. Implemented on HoloLens 2, SPMark was evaluated in a two-part study with 21 participants. In a controlled experiment, it improved completion rates and reduced perceived effort and frustration compared to a baseline. In open-ended exploration, users reported SPMark enabled precise, natural, and comfortable annotation. Our findings highlight the promise of smooth pursuit-enhanced gaze interaction for annotating augmented documents in tasks such as proofreading and structured note-taking.