Automatic Gesture Segmentation via Inertial Power Analysis

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing gestural interaction systems require users to manually specify the start and end of instrumented gestures by pressing a button, leading to visual load, constrained body movements, and potential musculo-skeletal disorders, as well as errors due to incorrect signal segmentation.

Innovation Solution

A method for automatic temporal segmentation of instrumented gestures using inertial navigation modules, which calculates instantaneous power values and estimates a gesture indicator based on energy variations over time, determining the start and end of gestures without user input, employing two estimators with different weightings to enhance accuracy and robustness.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a button is pressed during gesture execution to specify start and end points, then gesture segmentation accuracy is improved, but user convenience deteriorates due to visual load and constrained body movements

Engineering Contradiction:

Improvegesture segmentation accuracyVSAvoiduser convenience

Core Design Contradiction:

Measurement precisionVSEase of operation

Solution Approach 1:

The system automatically detects gesture start and end points by analyzing inertial sensor data without requiring user intervention. The processor autonomously segments the gesture signal by identifying characteristic patterns in acceleration, angular velocity, and magnetic field data, freeing the user from the burden of manual button pressing while maintaining accurate gesture recognition

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The manual mechanical interaction of button pressing is replaced by an automated sensor-based detection system. Inertial sensors (accelerometer, gyrometer, magnetometer) continuously monitor device movement and the processor algorithmically determines gesture boundaries based on physical motion characteristics, substituting the mechanical user action with an automated physical measurement system

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If a button is pressed continuously during gesture execution, then gesture boundaries are clearly defined, but musculo-skeletal disorders may occur due to constrained body movements

Engineering Contradiction:

Improvegesture boundary definitionVSAvoidmusculo-skeletal disorders

Core Design Contradiction:

ReliabilityVSObject-affected harmful factors

Solution Approach 1:

The system performs self-service by automatically detecting gesture boundaries through inertial sensor analysis, eliminating the need for continuous button pressing. The processor monitors physical movement patterns and autonomously identifies when gestures begin and end, ensuring reliable gesture boundary definition without requiring sustained finger pressure or constrained hand positioning

Inventive Principle:

Principle #25Self-service

3Measurement precision

If button pressing is required for gesture segmentation, then gesture start and end points are accurately identified, but errors occur due to premature finger removal

Engineering Contradiction:

Improvegesture timing identificationVSAvoidgesture recognition accuracy

Core Design Contradiction:

Measurement precisionVSReliability

Solution Approach 1:

The system replaces the mechanical button-pressing mechanism with automated inertial sensor-based detection. The processor analyzes continuous streams of acceleration, angular velocity, and magnetic field data to algorithmically determine gesture boundaries based on physical motion characteristics, eliminating errors caused by premature finger removal while maintaining accurate timing identification

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system continuously monitors inertial sensor data and provides real-time feedback on gesture progression. By analyzing the dynamic characteristics of device movement throughout the gesture execution, the processor can accurately identify start and end points based on physical motion patterns rather than arbitrary button presses, improving both precision and reliability

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS10078373B2Method of temporal segmentation of an instrumented gesture, associated device and terminal

Publication Date: 2018.09.18 ORANGE SA

AI summary

Temporally segmenting an instrumented gesture executed by a user with a terminal having an inertial navigation module, which measures a vector of inertial characteristics representative of movement of the terminal. Segmenting includes, at each current instant: calculating an instantaneous power value of the vector; estimating a gesture indicator based on variation between the instantaneous power value and a mean power value estimated over a preceding time window; determining a start of gesture at a first instant, when the estimated gesture indicator is greater than or equal to a first threshold during a time interval greater than or equal to a first interval; and determining an end of gesture at a second instant when, at the current instant, the estimated gesture indicator is less than or equal to a second threshold during a time interval greater than or equal to a second time interval.