Unified Classifier Model via Feature Mapping Across Devices

Solution Overview

Problem

Classifier models trained on one device platform, such as smartphones, often do not perform well on other platforms like tablets due to differences in form factors and sensor characteristics, making it impractical and costly to train individual classifiers for each device type.

Innovation Solution

A feature mapping function is generated to transform sensor data features from one device platform to match the statistical distribution of another, allowing a unified classifier model to be used across multiple devices, including smartphones and tablets, by applying statistical distribution distance minimization techniques like maximum mean discrepancy and principal component analysis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If classifiers are trained individually for each device platform to optimize recognition performance, then recognition accuracy is improved, but training cost and time consumption increase significantly

Engineering Contradiction:

Improverecognition accuracyVSAvoidtraining time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent introduces a feature mapping function as an intermediary transformation layer between sensor data from different device platforms and the unified classifier model. This mapping function adapts features from various devices (smartphones, tablets, wearables) to a common statistical distribution, enabling the classifier to process data from any device without retraining while maintaining recognition accuracy.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent creates a universal classifier model that can process sensor data from multiple device platforms (smartphones, tablets, wearables) through a single unified model. The feature mapping function enables this universality by transforming device-specific features into a platform-agnostic representation that the classifier can handle effectively across all device types.

Inventive Principle:

Principle #6Universality (Multi-functionality)

2Measurement precision

If classifiers are trained individually for each device platform, then recognition accuracy is improved, but manufacturing cost increases

Engineering Contradiction:

Improverecognition accuracyVSAvoidtraining cost

Core Design Contradiction:

Measurement precisionVSEase of manufacture

Solution Approach 1:

The patent creates a universal classifier model that can process sensor data from multiple device platforms (smartphones, tablets, wearables) through a single unified model. The feature mapping function enables this universality by transforming device-specific features into a platform-agnostic representation that the classifier can handle effectively across all device types.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent introduces a feature mapping function as an intermediary transformation layer between sensor data from different device platforms and the unified classifier model. This mapping function adapts features from various devices (smartphones, tablets, wearables) to a common statistical distribution, enabling the classifier to process data from any device without retraining while maintaining recognition accuracy.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Device complexity

If a unified classifier model is used across different device platforms, then development complexity is reduced, but recognition accuracy deteriorates due to platform differences

Engineering Contradiction:

Improveclassifier development complexityVSAvoidrecognition accuracy

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The patent introduces a feature mapping function as an intermediary transformation layer between sensor data from different device platforms and the unified classifier model. This mapping function adapts features from various devices (smartphones, tablets, wearables) to a common statistical distribution, enabling the classifier to process data from any device without retraining while maintaining recognition accuracy.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS11126897B2Unification of classifier models across device platforms

Publication Date: 2021.09.21 INTEL CORP

AI summary

Techniques are provided for unification of classifier models across device platforms of varying form factors and/or sensor calibrations. A methodology implementing the techniques according to an embodiment includes extracting classification features from data provided by sensors associated with a first device platform. The method also includes applying a feature mapping function to the extracted features. The feature mapping function is configured to transform the features such that the are suitable for use by a classifier model that is trained on data provided by sensors associated with a second device platform. The method further includes executing the classifier model on the transformed features to generate classifications, for example recognized activities associated with use of the first device. The feature mapping function is based on application of a statistical distribution distance minimization between a sampling of data provided by sensors of the first device and sensors of the second device.