Codec Performance Curve Fitting for Non-Overlapping Video Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing codec performance evaluation tools are limited by the requirement for monotonically increasing input data and overlapping data sets, leading to unstable and unreliable evaluations when these conditions are not met.
Innovation Solution
A method and apparatus for evaluating codec performance that allows for fitting non-monotonic data sets and non-overlapping intervals using polynomial functions with constraints to ensure accurate evaluation, including fitting curves with monotonic or convex properties.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing codec performance evaluation tools are used, then evaluation can be performed under standard conditions, but the tools become unstable and unreliable when data is non-monotonic or non-overlapping
Solution Approach 1:
The patent changes the mathematical parameters of curve fitting by using convex functions and monotonic function constraints instead of traditional polynomial fitting. This allows the evaluation tool to handle non-monotonic and non-overlapping data sets reliably by transforming the fitting approach to accommodate diverse data characteristics without compromising stability.
Solution Approach 2:
The patent introduces dynamic adaptability by allowing the curve fitting method to adjust based on the characteristics of the input data. The system dynamically selects appropriate fitting strategies (monotonic or convex constraints) depending on whether the data exhibits monotonic behavior, making the tool versatile across different codec performance scenarios while maintaining reliability.
2Measurement precision
If polynomial fitting is used for curve fitting, then mathematical simplicity is achieved, but the fitting becomes inaccurate for non-monotonic data sets
Solution Approach 1:
The patent applies local quality by imposing specific constraints (monotonic or convex) on different portions of the curve fitting process based on the local characteristics of the data. Instead of using a single global polynomial fitting approach, the method adapts the fitting constraints to match the local behavior of the codec performance data, thereby improving accuracy for non-monotonic data while maintaining manageable complexity.
3Stability of the object's composition
If monotonic constraint is applied to polynomial fitting, then fitting stability is improved, but the constraint becomes invalid for non-monotonic codec performance data
Solution Approach 1:
The patent makes the fitting constraint dynamic by selecting between monotonic constraints and convex constraints based on the characteristics of the input data. When data is monotonic, monotonic constraints are applied for stability; when data is non-monotonic, convex constraints are used instead. This dynamic adaptation maintains fitting stability across all data types while providing the versatility needed to handle diverse codec performance scenarios.
Data Source
AI summary
Methods, apparatus, and computer readable storage medium evaluating codec performance. One method includes obtaining m anchor data points each generated based on a respective anchor encoded video bitstream; obtaining n test data points each generated based on a respective encoded test video bitstream, n being an integer; fitting the m anchor data points with an anchor curve, the anchor curve being based on an anchor polynomial, wherein the anchor polynomial is monotonic in an x-axis range; fitting the n test data points with a test curve, the anchor curve being based on a test polynomial, wherein the test polynomial is monotonic in the x-axis range; and evaluating the test codec performance based on the anchor curve and the test curve, to obtain an evaluation result.