[0001] Techniques described herein provide an efficient technique for performing location-based tasks, which may be used by a variety of devices, including AR/VR devices. According to particular embodiments, an AR/VR device may selectively use localization techniques (e.g., SLAM) and machine learning (ML) to determine the location of the user, which in turn may be used to trigger downstream applications, such as personal assistant or recommendation services. Localization is used to determine the user’s specific 3D pose within the user’s environment, which would be useful for applications that need to know the user’s particular 3D pose. For example, in order to determine whether to render a virtual content that is anchored to the physical world, an AR/VR device would need to determine the user’s viewpoint relative to the virtual content in order to render the virtual content properly. In contrast to localization, ML-based location detection may not provide sufficient information about the user’s pose, but it would be able to detect where the user is generally located. For example, an ML model may be trained to detect which room the user is generally in. Such information may be all that is needed for applications to trigger room-based tasks, such as turning on lights in the room, sending reminders (e.g., a reminder to water the plants each time the user enters the kitchen), etc. While the ML-based location detection technique may not be able to provide accurate pose information, its operational cost is lower than localization techniques.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.

Drawings.pdf (11 kB)