Exponential QoE Model for Video Streaming Quality Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing quality of experience (QoE) models for video streaming services are inadequate as they lack intuitive parameters and robustness, failing to accurately predict user experience due to arbitrary parameter determination and inability to account for rebuffering events effectively.
Innovation Solution
An exponential QoE model is introduced, combining visual quality scores with rebuffering duration, using a linear regression model trained with subjective scores to provide a more intuitive and robust prediction of user experience, incorporating parameters like gain factor, non-linearity visual quality factor, and rebuffering factor.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If visual quality metrics are used to estimate quality of experience, then the quality of encoded video content is measured, but playback issues related to network bandwidth or throughput are not reflected
Solution Approach 1:
The patent combines visual quality metrics with rebuffering metrics into a unified QoE model. The exponential QoE model integrates both the quality of encoded video content and the impact of rebuffering events, creating a comprehensive assessment that neither metric could provide alone.
2Measurement precision
If existing QoE models with numerous parameters are used, then an attempt is made to predict quality of experience, but the parameters have no intuitive meanings and the models lack robustness
Solution Approach 1:
The patent extracts and focuses on the two most critical factors affecting QoE: visual quality and rebuffering duration. By eliminating unnecessary parameters and retaining only these essential elements with intuitive meanings, the model achieves both simplicity and predictive accuracy.
Solution Approach 2:
The patent transforms the QoE model into an exponential function with specific parameters that have clear interpretations. The visual quality exponent and rebuffering exponent are derived to reflect their respective impacts on user experience, making the model both mathematically rigorous and intuitively understandable.
3Reliability
If existing QoE models are used, then prediction of quality of experience is attempted, but the models are unable to accurately predict quality across a broad range of encoded video content and rebuffering events
Solution Approach 1:
The patent creates a universal QoE model that can accurately predict quality of experience across diverse video content and network conditions. The exponential formulation with configurable exponents allows the model to adapt to different scenarios while maintaining a consistent mathematical framework, enabling broad applicability without sacrificing accuracy.
Data Source
AI summary
In various embodiments, a quality of experience (QoE) prediction application computes a visual quality score associated with a stream of encoded video content. The QoE prediction application also determines a rebuffering duration associated with the stream of encoded video content. Subsequently, the QoE prediction application computes an overall QoE score associated with the stream of encoded video content based on the visual quality score, the rebuffering duration, and an exponential QoE model. The exponential QoE model is generated using a plurality of subjective QoE scores and a linear regression model. The overall QoE score indicates a quality level of a user experience when viewing reconstructed video content derived from the stream of encoded video content.