Head-Motion Emotion Detection with Physiological Confirmation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing emotion recognition technologies require specific equipment and stable methodologies, limiting their widespread use, and there is a need for methods that can reliably determine emotions using minimal hardware, such as a mobile device, especially when conversational data is not available.

Innovation Solution

A system that uses a machine learning model to analyze head movements and physiological parameters from a single wearable device, allowing emotion detection without conscious interaction and minimal hardware, with optional additional sensors for more complex versions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If full-body motion sensing input devices and stable methodologies are used for biological motion pattern analysis, then emotion detection accuracy is improved, but device complexity and hardware requirements increase

Engineering Contradiction:
Improveemotion detection accuracyVSAvoidhardware requirements
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts and focuses specifically on head motion patterns from the full-body motion data, isolating the most relevant feature for emotion detection. By using only head motion signals from simplified sensors rather than comprehensive full-body motion sensing, the system achieves effective emotion detection with reduced hardware complexity while maintaining acceptable accuracy.

Inventive Principle:
Principle #2Taking out (Extraction)

2Measurement precision

If multiple modalities and conversational data are used for emotion recognition, then recognition accuracy is improved, but ease of operation and user convenience deteriorate due to requiring conscious interaction

Engineering Contradiction:
Improverecognition accuracyVSAvoiduser convenience
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The system performs automatic emotion detection using unconscious head motion patterns without requiring the user to consciously provide data or interact with the device. The head-worn device continuously captures head motion signals and processes them through machine learning models to determine emotional states, enabling passive, hands-free operation that maintains accuracy while significantly improving user convenience.

Inventive Principle:
Principle #25Self-service

3Reliability

If specific equipment and stable methodologies are used for biological motion analysis, then reliability of emotion detection is improved, but adaptability and widespread use deteriorate

Engineering Contradiction:
Improveemotion detection reliabilityVSAvoidwidespread use
Core Design Contradiction:
ReliabilityVSAdaptability or versatility

Solution Approach 1:

The patent develops a universal emotion detection system using head motion patterns that can be applied across multiple contexts and platforms. By focusing on head motion—a universal human behavior captured by common head-worn devices like smartphones and earbuds—the system achieves both reliability through machine learning models and adaptability for widespread use in clinical practices, sales, media content analysis, and social robot interactions without requiring specialized equipment.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP4357886B1System, method and computer program for detecting and confirming an emotion of a user
Publication Date: 2025.07.02 ESSILOR INTERNATIONAL(COMPAGNIE GENERALE D OPTIQUE)
  • EP4357886B1 patent drawingFigure 1~2
  • EP4357886B1 patent drawingFigure 3~4
  • EP4357886B1 patent drawingFigure 5

AI summary

The disclosure relates to a system configured to provide a first signal representing at least movements of the head of a user to a machine learning model and consequently obtain an output representing an emotion of the user, the machine learning model having been trained beforehand using a database of model signals representing movements of a head of a model user and associated with at least an emotion of the model user, and process a second signal representing a physiological parameter of the user to confirm the emotion of the user. The disclosure further relates to a corresponding method and a corresponding computer program.