Linear Fusion Model for Virtual Desktop Visual Perception
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image quality metrics fail to accurately measure visual perception in virtual desktop environments, leading to subjective evaluation challenges and inefficient resource allocation, as they do not effectively correlate with human subjective quality ratings due to the complexity of desktop screen imagery.
Innovation Solution
A linear fusion model combining peak signal to noise ratio (PSNR), structural similarity (SSIM), and feature similarity (FSIM) scores is trained using a benchmark database of distorted virtual desktop images and subjective human ratings to provide an objective, accurate visual perception quality metric.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional image quality metrics are used to measure visual perception in virtual desktop environments, then measurement simplicity is maintained, but measurement precision deteriorates because these metrics fail to accurately correlate with human subjective quality ratings
Solution Approach 1:
The patent combines multiple image quality metrics (PSNR, SSIM, FSIM) into a single composite quality score through a linear fusion model. This merging approach allows the system to capture different aspects of image quality (signal-to-noise ratio, structural similarity, feature similarity) and integrate them into one comprehensive measurement that better correlates with human subjective perception, thereby improving measurement precision without requiring multiple separate evaluation systems
Solution Approach 2:
The patent transforms the measurement approach by changing from single-parameter metrics to a multi-parameter fusion model. By adjusting the weighting parameters in the linear fusion model (α, β, γ), the system can optimize the contribution of each metric component to better match human perception characteristics. This parameter optimization enables accurate visual perception quality assessment while maintaining computational efficiency
2Measurement precision
If subjective evaluation methods are used to assess visual perception quality, then alignment with human perception is achieved, but ease of operation deteriorates due to the need for human participants and subjective judgment variability
Solution Approach 1:
The patent creates an objective computational model that copies and simulates human subjective perception characteristics. Instead of directly querying human subjects, the system uses the linear fusion model trained on subjective ratings to replicate human judgment patterns. This allows the system to achieve high correlation with human perception while eliminating the operational complexities of conducting actual subjective tests, including participant recruitment, test administration, and result aggregation
Solution Approach 2:
The patent replaces the mechanical system of human subjective evaluation with an automated computational evaluation system. The linear fusion model acts as a substitute that performs the same function (assessing visual perception quality) without requiring human participants. This substitution maintains the accuracy benefits of subjective evaluation while eliminating its operational drawbacks, enabling automated, consistent, and scalable quality assessment
3Measurement precision
If lossless quality (100% quality) is provided in virtual desktop protocols, then visual perception quality is maximized, but loss of energy increases due to higher bandwidth consumption
Solution Approach 1:
The patent enables partial quality action by allowing the system to deliver image quality that is sufficient for the required visual perception level without necessarily achieving complete lossless quality. The linear fusion model determines the actual quality level needed, and the system can operate at this optimized level rather than always providing maximum quality. This partial action approach maintains acceptable visual perception while reducing bandwidth consumption when full lossless quality is not necessary
Solution Approach 2:
The patent introduces dynamic quality adjustment capabilities through the linear fusion model. The system can adaptively adjust the quality level based on current network conditions, user preferences, and the specific visual content being displayed. By dynamically optimizing the balance between quality and bandwidth usage, the system maintains high visual perception quality when needed while reducing energy loss during periods when lower quality is acceptable, enabling flexible resource allocation
Data Source
AI summary
Techniques are described for improving the measurement of visual perception of graphical user interface (GUI) information remoted to client devices in virtual desktop environments, such as VDI and DAAS. An objective image quality measurement of remoted virtual desktop interfaces is computed, that is more accurate and more closely aligned with subjective user perception. The visual quality metric is computed using a linear fusion model that combines a peak signal to noise ratio (PSNR) score of the distorted image, a structural similarity (SSIM) score of the distorted image and a feature similarity (FSIM) score of the distorted image. Prior to using the model to compute the quantitative visual perception metric, the linear fusion model is trained by using a benchmark test database of reference images (e.g., virtual desktop interface images), distorted versions of those images and subjective human visual perception quality ratings associated with each distorted version.