Head Pose Tracking via Correlated Motion Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing spatial audio platforms face challenges in maintaining accurate head pose tracking when the source device and headset are free to move relative to each other, leading to issues such as audio swiveling off-center during movements like turning in a bus or plane.
Innovation Solution
The implementation of correlated motion detection, which involves obtaining motion data from both the source device and the headset, determining correlation measures, updating the motion tracking state, and initiating motion tracking accordingly. This allows the system to transition between single IMU and two IMU tracking states based on the correlation between the source device and headset motions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single IMU is used in the headset for head pose tracking, then the device complexity is reduced, but the audio swivels off-center when the source device and headset move relative to each other (e.g., during bus or plane turns)
Solution Approach 1:
The patent combines data from two IMUs (one in the headset and one in the source device) to perform head pose tracking. By merging the motion data from both devices and computing correlation measures, the system can distinguish between head movements and source device movements, preventing audio swiveling while maintaining reasonable system complexity.
Solution Approach 2:
The patent introduces correlation computation as an intermediary process between the two IMUs. The correlation measure acts as a mediator to determine whether observed motion is due to head movement or source device movement, enabling the system to selectively apply tracking corrections only when appropriate.
2Reliability
If motion tracking is always performed relative to source device rotation, then audio remains centered during correlated motion, but the system fails to respond to actual head pose changes during uncorrelated motion
Solution Approach 1:
The patent dynamically adjusts the motion tracking state based on real-time correlation measures. The system transitions between two states: relative motion tracking (when correlation is high) and absolute motion tracking (when correlation is low). This dynamic adaptation ensures accurate audio positioning regardless of whether the source device or headset is moving.
Solution Approach 2:
The patent uses correlation measures as feedback to continuously monitor the relationship between source device and headset motion. This feedback loop enables the system to detect when the source device is rotating independently of the headset and adjust tracking accordingly, maintaining audio accuracy in varying conditions.
3Stability of the object's composition
If correlation measures are computed using large time windows, then the detection is more stable, but the response time to motion changes is delayed
Solution Approach 1:
The patent segments the correlation computation into multiple time windows of different durations. By computing correlation measures across multiple time scales, the system achieves both stability (through longer windows) and responsiveness (through shorter windows), allowing it to detect motion changes quickly while maintaining reliable correlation assessment.
Data Source
AI summary
Embodiments are disclosed for head tracking state detection based on correlated motion of a source device and a headset communicatively coupled to the source device. In an embodiment, a method comprises: obtaining, using one or more processors of a source device, source device motion data from a source device and headset motion data from a headset; determining, using the one or more processors, correlation measures using the source device motion data and the headset motion data; updating, using the one or more processors, a motion tracking state based on the determined correlation measures; and initiating head pose tracking in accordance with the updated motion tracking state.


