Adaptive Multimedia Quality Parameter Generation Using Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Determining an optimal combination of image and sound quality parameters from an enormous number of possibilities is difficult for skilled technicians, and these parameters often vary with different multimedia data, leading to inconsistent quality outcomes.
Innovation Solution
A learning device that performs quantitative evaluation on multimedia data, selects suitable parameters through machine learning, and generates models to automatically determine optimal image and sound quality parameters for processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a skilled technician manually determines image quality parameters, then parameter selection can be done with expertise, but the process requires enormous time and labor due to the vast number of combinations
Solution Approach 1:
The patent replaces the manual mechanical process of parameter selection by skilled technicians with an automated machine learning system. The learning device uses neural networks to automatically determine optimal image quality parameters from vast combinations, eliminating the need for human experts to manually evaluate each possibility while maintaining high selection accuracy.
Solution Approach 2:
The system enables self-service parameter determination where the machine learning model automatically selects optimal parameters without requiring skilled technician intervention. The learning device evaluates and determines parameters independently based on input images, making the process accessible without expert labor.
2Ease of manufacture
If a fixed image quality parameter is applied to multiple images, then the process is simple and consistent, but the parameter may not be optimal for each specific image
Solution Approach 1:
The patent transitions from static fixed parameters to dynamic parameter selection. The machine learning model analyzes each image's characteristics and dynamically determines the most suitable image quality parameters for that specific image, allowing the system to adapt to varying image content while maintaining automated processing simplicity.
Solution Approach 2:
The system automatically changes parameters based on image characteristics. The learning device determines different image quality parameters for different images rather than using a fixed set, enabling optimization for each specific image while maintaining a simple automated workflow through machine learning-driven parameter adaptation.
3Reliability
If manual parameter determination is used, then expert judgment can be applied, but the process cannot scale efficiently with large volumes of multimedia data
Solution Approach 1:
The patent replaces manual expert judgment with automated machine learning inference. The learning device uses trained neural networks to rapidly determine optimal parameters for large volumes of multimedia data, maintaining reliable parameter selection through learned patterns while achieving high processing throughput that scales with data volume.
Solution Approach 2:
The system enables continuous automated parameter determination without interruption. The machine learning model can process large volumes of data continuously, maintaining consistent and reliable parameter selection across all inputs while achieving high throughput through uninterrupted automated operation.
Data Source
AI summary
In a learning device (10), a first evaluation unit (14) performs quantitative evaluation on a plurality of pieces of image data and thereby acquires a plurality of first evaluation results for each of the plurality of pieces of image data, a teacher data generation unit (15) selects a second parameter from among a plurality of first parameters having values different from each other on the basis of the plurality of first evaluation results and generates a first set of teacher data including the selected second parameter, and a first machine learning unit (17) performs machine learning using first sets of teacher data and thereby generates a learned model that outputs a third parameter used for processing of image data of a processing target.