Image Labeling Interface Using Clustering for Training Datasets

Solution Overview

Problem

The process of creating training datasets for image recognition algorithms is time-consuming and expensive, particularly when specialized human review is required for labeling images, which can be a bottleneck in the training process.

Innovation Solution

A computer system employs machine learning algorithms to derive features from images, reduce dimensionality, cluster datapoints, and generate visual representations for user input, facilitating efficient labeling of images for training datasets.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If human reviewers manually label each image in the training dataset, then the accuracy of the training dataset is improved, but the time and cost required for preparation increases significantly

Engineering Contradiction:

Improveaccuracy of training datasetVSAvoidtime required for user review process

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary actions by using machine learning algorithms to automatically derive features from images, reduce dimensionality, and generate cluster predictions before human review. This preprocessing organizes the data and provides initial labels, reducing the subsequent manual work required while maintaining accuracy.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system introduces an intermediary machine learning model that acts as a mediator between raw images and final labeled training data. This intermediary automatically generates feature extractions, dimensionality reductions, and cluster predictions, which then guide or pre-populate the human review process, reducing time and cost.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Measurement precision

If human reviewers manually label each image in the training dataset, then the accuracy of the training dataset is improved, but the cost of preparation increases significantly

Engineering Contradiction:

Improveaccuracy of training datasetVSAvoidcost of preparing training dataset

Core Design Contradiction:

Measurement precisionVSEase of manufacture

Solution Approach 1:

The system performs preliminary actions by using machine learning algorithms to automatically derive features from images, reduce dimensionality, and generate cluster predictions before human review. This preprocessing organizes the data and provides initial labels, reducing the subsequent manual work required while maintaining accuracy.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system introduces an intermediary machine learning model that acts as a mediator between raw images and final labeled training data. This intermediary automatically generates feature extractions, dimensionality reductions, and cluster predictions, which then guide or pre-populate the human review process, reducing time and cost.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Reliability

If a large number of images are included in the training dataset, then the quality of image recognition algorithm improves, but the time and resources required for labeling increases

Engineering Contradiction:

Improvequality of image recognition algorithmVSAvoidefficiency of training dataset preparation

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system performs preliminary actions by using machine learning algorithms to automatically derive features from images, reduce dimensionality, and generate cluster predictions before human review. This preprocessing organizes the data and provides initial labels, reducing the subsequent manual work required while maintaining accuracy.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system introduces an intermediary machine learning model that acts as a mediator between raw images and final labeled training data. This intermediary automatically generates feature extractions, dimensionality reductions, and cluster predictions, which then guide or pre-populate the human review process, reducing time and cost.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentEP4032020B1Using machine learning algorithms to prepare training datasets

Publication Date: 2026.01.28 LUMINEX CORP

Figure 1
Figure 2
Figure 3

AI summary

A computer system is configured to receive a dataset of image-derived features for a plurality of images, reduce the dimensionality of this dataset, identify clusters within the dimensionally-reduced dataset, and generate a visual representation of the datapoint of the dimensionally-reduced dataset as icons grouped by cluster. User input is received to apply user classification labels to the images for inclusion in a training dataset. A user interface is useable to present information to the user and receive information from the user to facilitate the application of user classification labels.