Visual Quality Assessment Using Game State Data Cues
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing visual quality assessment models for multimedia, particularly gaming content, fail to differentiate between intentional and unintentional image degradation in video games, leading to inaccurate quality assessments.
Innovation Solution
A visual quality assessment method and system that incorporates supplementary inputs from game state data such as texture information, polygon count, game engine settings, post-processing effects, and input device signals to adjust quality assessment models like PSNR, SSIM, VMAF, and machine learning algorithms, accounting for intentional image effects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional VQA models (PSNR, SSIM, VMAF) are used for gaming content, then the assessment process is simple and fast, but the accuracy is poor because they cannot differentiate between intentional and unintentional image degradation
Solution Approach 1:
The assessment system is segmented into multiple specialized components: a game state data acquisition module that collects supplementary information about intentional effects, a modification module that adjusts VQA calculations based on this data, and integration with traditional VQA models. This segmentation allows the system to maintain simplicity where possible while adding complexity only where needed to differentiate intentional from unintentional degradation.
Solution Approach 2:
Game state data acts as an intermediary element that mediates between the image content and the VQA assessment. This supplementary information about intentional effects (motion blur, color desaturation, etc.) allows the assessment system to interpret visual artifacts correctly without requiring complete redesign of the underlying VQA models.
2Measurement precision
If traditional VQA models are applied to gaming content, then computational resources are conserved, but the assessment results are inaccurate due to inability to account for intentional image effects
Solution Approach 1:
Game state data is acquired and processed in advance before the actual VQA assessment occurs. This preliminary action prepares the necessary information about intentional effects upfront, allowing the main VQA computation to proceed efficiently with pre-processed guidance rather than requiring complex real-time analysis during the assessment phase.
Solution Approach 2:
The system applies modified assessment calculations selectively - only adjusting VQA computations where game state data indicates the presence of intentional effects. This local quality approach ensures that additional computational resources are consumed only when and where needed to account for intentional image degradation, rather than uniformly across all assessment scenarios.
Data Source
AI summary
A visual quality assessment (VQA) method includes the steps of, for all or part of an image, obtaining data indicative of an intended graphical state of the image or corresponding part thereof, from a device generating the image; and performing a VQA based on the image or corresponding part thereof, wherein at least one calculation step of the VQA is modified responsive to the obtained data.