Abstract
Users interacting with head-wearable augmented reality (AR) devices frequently experience virtual content clutter, such as incoming notifications, application windows, or video feeds overlaying their primary field of view. When a user engages in a face-to-face social interaction while wearing head-wearable devices, this virtual content can inadvertently obstruct the user's direct line of sight to the conversation partner's face and upper body. This obstruction disrupts natural eye contact and non-verbal communication cues, creating social disconnect and distraction. Furthermore, relying on manual hand gestures to adjust or reposition virtual content is cumbersome and interrupts conversational flow.
The disclosed technology describes a conversation-aware display management framework for head-wearable AR devices that automatically repositions virtual content based on social context. An integrated contextual fusion engine evaluates real-time data streams from multiple embedded sensors, performing sensor fusion across gaze tracking from user-facing cameras, facial landmark and upper body recognition from forward-facing cameras, vocal speech activity detection from audio sensors, and distance estimation from depth sensors. The contextual fusion engine utilizes the multiple data streams to determine if the wearer is actively involved in an in-person conversation with a conversation partner. Upon confirming that the wearer is engaged in a face-to-face conversation within a typical conversational distance, the system dynamically calculates a spatial exclusion zone identifying the location of the conversation partner's face and, in some cases, upper body.
When active virtual content (e.g., virtual windows) intersects the established exclusion zone, the AR operating environment executes an automated user interface transition to move the virtual content to coordinates at the display periphery. For example, the intersecting virtual content may be shifted toward the display edges, scaled down, or reduced in opacity to ensure the conversation partner remains visible. Following the conclusion of the conversation, the virtual content automatically returns to its original display coordinates, while high-priority emergency alerts bypass suppression to preserve user safety. This architecture provides unobtrusive, context-aware digital assistance that preserves interpersonal human connection and reduces cognitive load without completely hiding notifications.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Singh, Ashish and Öztireli, Ahmet Cengiz, "Method and System for Dynamically Repositioning Virtual Content in Augmented Reality Systems During Conversations", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/12251