Part-Based Video Quality Estimation for Web Conferences
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video quality estimation techniques for video distribution services fail to accurately estimate the quality of Web conference videos due to differences in screen layouts and video characteristics, such as face and material videos, which are combined and reduced in size, leading to inaccurate quality assessment.
Innovation Solution
A video quality estimation apparatus that estimates video quality based on the quality of each part forming the video, considering the layout resolution, encoding information, and the influence of reduction and integration of parts, using a model that reflects the unique characteristics of face and material videos.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional video quality estimation techniques for video distribution services are applied to Web conference videos, then the estimation process is simple, but the estimation accuracy is low due to differences in screen layouts and video characteristics
Solution Approach 1:
The video is divided into multiple parts (e.g., face video, material video, background video) that are arranged in a screen layout. Each part is estimated separately using the video quality estimation model, and then the results are integrated based on the screen layout to obtain the overall video quality. This segmentation approach allows the model to account for the unique characteristics of each video part while maintaining a systematic estimation process.
2Measurement precision
If the video quality of each part is considered along with layout resolution, then the estimation accuracy is improved, but the computational complexity increases
Solution Approach 1:
The video quality estimation is performed locally for each part of the video (e.g., face video, material video) based on its specific characteristics and encoding information. The local quality estimates are then combined using the screen layout information to derive the overall video quality. This approach allows the system to focus computational resources on estimating the quality of each individual part rather than processing the entire video as a single unit.
3Ease of manufacture
If a generic video quality estimation model is used, then the model is easy to implement, but it cannot accurately reflect the unique characteristics of Web conference videos with screen layouts
Solution Approach 1:
The video quality estimation model is designed to be universal by accepting video data with screen layouts in which parts are arranged as input. The model can process various types of videos (e.g., Web conference videos, video distribution services) by adapting to their specific characteristics through the screen layout information. This universality allows the same model to be applied across different video types while maintaining accuracy for each specific case.
Data Source
AI summary
In a video quality estimation apparatus for estimating a video quality of a video having a screen layout in which parts are arranged, the video quality of the video is estimated based on a video quality of each of the parts forming the video and a layout resolution of each of the parts on the screen layout.