Adaptive Hidden Area Mask for HMD GPU Load Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

High-resolution and high-framerate requirements in head-mounted displays (HMDs) for virtual reality experiences lead to increased graphical processing unit (GPU) loading, making HMDs costly for users.

Innovation Solution

A method and system for adaptively adjusting the hidden area mask in HMDs to reduce GPU loading by dynamically enlarging the mask based on the wearer's eyesight, using an image capturing device and processor to track eye movement and adjust the visible area, thereby processing less image data without compromising the VR experience.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If high-resolution and high-framerate displays are used in HMDs to provide VR experiences, then the quality of VR experience is improved, but the GPU loading increases and device cost increases

Engineering Contradiction:
Improvedisplay resolutionVSAvoidGPU loading
Core Design Contradiction:
Manufacturing precisionVSPower

Solution Approach 1:

The patent applies partial action by rendering high-resolution images only in the foveal region (central vision) where users actually look, while rendering lower-resolution images in peripheral regions. This reduces the total number of pixels that need to be processed by the GPU, thereby reducing GPU loading and device cost while maintaining perceived VR experience quality.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The patent implements local quality by varying the rendering resolution across different regions of the visual field. The foveal region receives high-resolution rendering matching the high-resolution display capabilities, while peripheral regions receive lower-resolution rendering. This localized quality adjustment reduces overall computational load on the GPU while preserving visual quality where it matters most to the user experience.

Inventive Principle:
Principle #3Local quality

2Speed

If high-framerate displays are used in HMDs to provide VR experiences, then the quality of VR experience is improved, but the GPU loading increases and device cost increases

Engineering Contradiction:
ImproveframerateVSAvoidGPU loading
Core Design Contradiction:
SpeedVSPower

Solution Approach 1:

The patent reduces the computational burden for achieving high framerates by applying foveated rendering. Instead of rendering all pixels at full resolution at high framerate, only the foveal region is rendered at high resolution and high framerate, while peripheral regions are rendered at lower resolution. This partial action approach maintains the high-framerate VR experience in the critical foveal area while reducing overall GPU loading.

Inventive Principle:
Principle #16Partial or excessive action

3Power

If the hidden area mask is enlarged to reduce the visible area, then the GPU loading is reduced, but the visible area for the user is reduced

Engineering Contradiction:
ImproveGPU loadingVSAvoidvisible area
Core Design Contradiction:
PowerVSArea of stationary object

Solution Approach 1:

The patent employs dynamics by making the mask region adaptive and changeable based on user eye tracking data. Rather than using a fixed enlarged mask that permanently reduces visible area, the system dynamically adjusts the mask boundaries to follow the user's gaze, ensuring that the high-resolution rendered content is always visible in the unmasked region. This dynamic adaptation maintains both reduced GPU loading and adequate visible area for the user experience.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP3407167B1Head-mounted display system, method for adaptively adjusting hidden area mask, and computer program product
Publication Date: 2023.11.22 HTC CORP
  • EP3407167B1 patent drawingFigure 1
  • EP3407167B1 patent drawingFigure 2
  • EP3407167B1 patent drawingFigure 3

AI summary

The disclosure proposes a head-mounted display system, a method for adaptively adjusting a hidden area mask, and a computer program product. The method includes: displaying a plurality of first content to an eye of a wearer of the head mounted display system, wherein the first content is masked by a first hidden area mask; capturing a plurality of first eye images of the eye and obtaining an eyesight of the eye based on the first eye images; adjusting the first hidden area mask to be a second hidden area mask based on the eyesight of the eye; and displaying a second content to the eye, wherein the second content is masked by the second hidden area mask.