VR Sensor Data Filtering via Foreground Masking and Outlier Removal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Sensor data in virtual reality environments is often unreliable due to noise and occlusion, leading to improper user interaction with VR applications.
Innovation Solution
A method involving a host device with a processor that filters sensor data using statistical models and masks to remove outliers, and employs a prediction model to fill in missing data based on user images, enhancing data reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If sensor data is collected in noisy environments, then more interaction data is available, but the reliability of the sensor data deteriorates due to noise and occlusion
Solution Approach 1:
The patent extracts and removes outlier data points from the sensor data stream using statistical methods (e.g., standard deviation thresholding, interquartile range). This separates the reliable data from the unreliable noisy data, keeping only the valid portions for VR application control.
Solution Approach 2:
The patent introduces an intermediary data processing layer between the sensor and the VR application. This intermediary performs statistical analysis, outlier detection, and data imputation to clean and complete the sensor data before it reaches the application, acting as a buffer against environmental noise.
Solution Approach 3:
The patent performs preliminary data cleaning and validation operations on sensor data before it is used for VR interaction. By pre-processing the data to remove outliers and fill missing values in advance, the system ensures that only reliable data reaches the application layer.
2Reliability
If statistical filtering is applied to remove outliers, then data reliability improves, but data loss increases due to removal of potentially valid data points
Solution Approach 1:
The patent uses feedback mechanisms where the system continuously monitors the distribution and characteristics of sensor data, adjusting the outlier detection thresholds dynamically. This feedback loop ensures that the filtering criteria adapt to changing environmental conditions, reducing false positives and data loss.
Solution Approach 2:
Before applying aggressive filtering, the system performs preliminary analysis of the sensor data distribution to establish appropriate thresholds. This preliminary characterization of normal vs. abnormal data patterns helps preserve valid data points that might otherwise be incorrectly identified as outliers.
3Loss of information
If prediction models are used to fill missing data, then data completeness improves, but computational complexity increases
Solution Approach 1:
The patent employs multiple imputation methods with different levels of complexity (e.g., mean imputation, regression imputation, iterative imputation) and allows dynamic selection based on the severity of data loss and available computational resources. This parameter adjustment balances completeness against computational load.
Solution Approach 2:
The patent applies prediction models selectively rather than universally - using simpler imputation methods for minor data gaps and more sophisticated models only when necessary. This partial application of complex algorithms reduces overall computational complexity while still achieving adequate data completeness.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The embodiments of the disclosure provide a method for managing sensor data, a host, and a computer readable storage circuit. The method includes: obtaining a plurality of first screen frames of a first application and obtaining a first sensor data associated with first screen frames from a first sensor; determining a first foreground part within each of first screen frames; determining a first mask associated with the first sensor based on the first foreground part within each of first screen frames and the first sensor data associated with first screen frames; obtaining a plurality of second screen frames of the first application and obtaining a second sensor data associated with second screen frames from the first sensor; managing the second sensor data associated with second screen frames by using the first mask; and interacting with the first application with the managed second sensor data.