GhostUI: Unveiling Hidden Interactions in Mobile UI

TL;DR

GhostUI introduces a comprehensive dataset to improve the detection of hidden, gesture-based interactions in mobile apps, enhancing the capabilities of vision language models for better automation and user experience understanding.

Contribution

The paper presents GhostUI, a novel dataset with rich annotations that enables vision language models to better detect concealed interactions in mobile UIs, advancing mobile automation research.

Findings

Models fine-tuned on GhostUI outperform baselines in hidden interaction detection.

GhostUI improves post-interaction screen prediction accuracy.

Enhanced recognition of implicit gestures in mobile interfaces.

Abstract

Modern mobile applications rely on hidden interactions--gestures without visual cues like long presses and swipes--to provide functionality without cluttering interfaces. While experienced users may discover these interactions through prior use or onboarding tutorials, their implicit nature makes them difficult for most users to uncover. Similarly, mobile agents--systems designed to automate tasks on mobile user interfaces, powered by vision language models (VLMs)--struggle to detect veiled interactions or determine actions for completing tasks. To address this challenge, we present GhostUI, a new dataset designed to enable the detection of hidden interactions in mobile applications. GhostUI provides before-and-after screenshots, simplified view hierarchies, gesture metadata, and task descriptions, allowing VLMs to better recognize concealed gestures and anticipate post-interaction states. Quantitative evaluations with VLMs show that models fine-tuned on GhostUI outperform baseline VLMs, particularly in predicting hidden interactions and inferring post-interaction screens, underscoring GhostUI's potential as a foundation for advancing mobile task automation.