IP Library Granted Patent US 12704901
Granted Patent B2
US 12704901 · App. 19/058,864 · Granted Aug 11, 2026

Bending-assisted calibration for extended reality tracking

Inventors: Thomas Faeulhammer (Vienna, AT); Matthias Kalkgruber (Vienna, AT); Thomas Muttenthaler (Vienna, AT); Tiago Miguel Pereira Torres (Vienna, AT); Daniel Wolf (Mödling, AT)
Assignee: Snap Inc.
G06F3/012A61B5/16G06T17/00
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12704901
App. No.
19/058,864
Granted
Aug 11, 2026
Kind
B2
Abstract

Bending data is used to facilitate tracking operations of an extended reality (XR) device, such as hand tracking or other object tracking operations. The XR device obtains bending data indicative of bending of the XR device to accommodate a body part of a user wearing the XR device. The XR device determines, based on the bending data, whether to use previously identified biometric data in a tracking operation. A mode of the XR device is selected based on this determination. The XR device performs the tracking operation based on the selected mode. The selected mode may be a first mode in which the previously identified biometric data is used in the tracking operation or a second mode which does not apply previously identified biometric data in the tracking operation.

Claims (59)

1 . A method comprising:

obtaining bending data indicative of bending of an extended reality (XR) device to accommodate a body part of a user wearing the XR device;

determining, based on the bending data, whether to use previously identified biometric data in a tracking operation of the XR device;

selecting a mode of the XR device, the selected mode being a first mode in which the previously identified biometric data is applied in the tracking operation or a second mode in which previous biometric data is not applied in the tracking operation; and

performing, by the XR device, the tracking operation based on the selected mode.

2 . The method of claim 1 , wherein the previously identified biometric data comprises a hand feature estimate generated during a previous user session.

3 . The method of claim 1 , wherein obtaining the bending data comprises:

analyzing images from a plurality of cameras of the XR device having overlapping fields of view to estimate the bending of the XR device.

4 . The method of claim 3 , wherein analyzing the images comprises:

generating a depth map based on matched features between views from the plurality of cameras; and

analyzing the depth map to estimate the bending of the XR device.

5 . The method of claim 1 , wherein performing the tracking operation comprises:

implementing a detection phase identifying an object of interest; and

implementing a tracking phase estimating a pose of the object over time.

6 . The method of claim 5 , wherein the detection phase comprises:

detecting the object of interest in a camera field of view;

detecting landmarks on the object of interest; and

finding the landmarks across a plurality of camera images.

7 . The method of claim 5 , wherein the tracking phase comprises:

estimating a location of the object of interest based on at least one of object appearance, object motion, or object landmarks; and

updating the estimated pose of the object of interest based on the estimated location.

8 . The method of claim 1 , wherein performing the tracking operation in the first mode comprises:

loading the previously identified biometric data, the previously identified biometric data comprising a hand feature estimate;

computing a three-dimensional model of a hand based on the hand feature estimate; and

constraining parameters for tracking of the hand based on the hand feature estimate.

9 . The method of claim 8 , wherein the hand feature estimate is based on one or more bones of the hand.

10 . The method of claim 1 , wherein determining, based on the bending data, whether to use the previously identified biometric data in the tracking operation of the XR device comprises:

comparing the bending data to previously stored bending data associated with the previously identified biometric data; and

determining that the bending matches the previously stored bending data when the bending data is within an acceptable range of the previously stored bending data.

11 . The method of claim 10 , wherein the acceptable range is defined by a threshold difference between the bending data and the previously stored bending data.

12 . The method of claim 1 , wherein determining, based on the bending data, whether to use the previously identified biometric data in the tracking operation of the XR device comprises:

processing the bending data to generate a prediction for biometric data; and

comparing the prediction with the previously identified biometric data.

13 . The method of claim 1 , wherein the selected mode is the first mode, and performing the tracking operation in the first mode comprises:

loading the previously identified biometric data; and

initializing the tracking operation using the loaded previously identified biometric data.

14 . The method of claim 1 , wherein the selected mode is the second mode, and performing the tracking operation in the second mode comprises:

performing a calibration operation to obtain new biometric data;

storing the new biometric data; and

performing the tracking operation using the new biometric data.

15 . The method of claim 1 , wherein the XR device comprises a head-wearable device including a flexible frame configured to bend to accommodate a head of the user.

16 . The method of claim 1 , wherein obtaining the bending data comprises using one or more strain gauge sensors to detect deformation.

17 . The method of claim 1 , wherein obtaining the bending data comprises:

measuring a strain of the XR device;

detecting changes in the strain when the XR device bends to accommodate a head of the user wearing the XR device; and

generating the bending data based on the changes in the strain.

18 . The method of claim 1 , wherein the tracking operation comprises tracking hand gestures for user input to the XR device.

19 . A system comprising:

at least one processor; and

at least one memory storing instructions that, when executed by the at least one processor, configure the system to perform operations comprising:

obtaining bending data indicative of bending of an extended reality (XR) device to accommodate a body part of a user wearing the XR device;

determining, based on the bending data, whether to use previously identified biometric data in a tracking operation of the XR device;

selecting a mode of the XR device, the selected mode being a first mode in which the previously identified biometric data is applied in the tracking operation or a second mode in which previous biometric data is not applied in the tracking operation; and

performing, by the XR device, the tracking operation based on the selected mode.

20 . One or more non-transitory computer-readable storage media, the computer-readable storage media including instructions that when executed by at least one processor, cause the at least one processor to perform operations comprising:

obtaining bending data indicative of bending of an extended reality (XR) device to accommodate a body part of a user wearing the XR device;

determining, based on the bending data, whether to use previously identified biometric data in a tracking operation of the XR device;

selecting a mode of the XR device, the selected mode being a first mode in which the previously identified biometric data is applied in the tracking operation or a second mode in which previous biometric data is not applied in the tracking operation; and

performing, by the XR device, the tracking operation based on the selected mode.