Optical Map Quality Evaluation and Augmentation Network

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing image processing systems on mobile devices face challenges in objectively evaluating the quality of optical maps, such as depth and disparity maps, due to subjective assessment methods, and data augmentation techniques often result in reduced accuracy of machine learning models when using brute-force random augmentation, failing to model real-world data variations effectively.

Innovation Solution

The system employs a method to quantify optical map quality by generating multiple maps using different algorithms and scoring them against ground truth maps, and uses an augmentation network to generate training data that mimics real-world image variations, improving the accuracy of machine learning models by providing more realistic data for training.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If brute-force random data augmentation is applied to training data, then the quantity of training data increases, but the accuracy of the machine learning model decreases

Engineering Contradiction:

Improvequantity of training dataVSAvoidaccuracy of machine learning model

Core Design Contradiction:

Quantity of substanceVSMeasurement precision

Solution Approach 1:

The patent transforms the training data by applying parameter changes that mimic real-world variations (photometric transformations, geometric transformations, noise patterns) rather than random changes. This allows the model to learn from diverse data patterns while maintaining accuracy, resolving the contradiction between data quantity and model accuracy.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The system uses feedback from real-world performance evaluation to refine the augmentation process. By evaluating how well augmented data represents real-world scenarios and adjusting the augmentation parameters accordingly, the system maintains model accuracy while expanding training data effectiveness.

Inventive Principle:

Principle #23Feedback

2Ease of operation

If subjective assessment methods are used to evaluate optical map quality, then the evaluation process is simple, but the objectivity and reliability of quality assessment deteriorates

Engineering Contradiction:

Improvesimplicity of evaluation processVSAvoidobjectivity of quality assessment

Core Design Contradiction:

Ease of operationVSReliability

Solution Approach 1:

The patent replaces subjective human assessment with an automated computational evaluation system that uses objective metrics (such as correlation coefficients, root mean square error, or other quantitative measures) to assess optical map quality. This substitution maintains ease of operation while dramatically improving objectivity and reliability of the assessment.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS11847771B2Systems and methods for quantitative evaluation of optical map quality and for data augmentation automation

Publication Date: 2023.12.19 SAMSUNG ELECTRONICS CO LTD

AI summary

A method includes obtaining at least one image and a ground truth map associated with the at least one image. The method also includes generating multiple optical maps using multiple algorithms and the at least one image. The method further includes, for each algorithm, identifying at least one score for the algorithm using one or more of the optical maps generated using the algorithm and the ground truth map. The ground truth map identifies one or more boundaries associated with one or more foreground objects in the at least one image. The scores identify how well the optical maps generated using the algorithms separate the one or more foreground objects from a background in the at least one image.