Gesture Recognition Using Multiple Sensors and Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing electronic devices struggle to accurately detect user gestures without relying on touch input, particularly in scenarios where touch input is not feasible or desirable.
Innovation Solution
The use of multiple sensors, including biosignal sensors and non-biosignal sensors, to provide input to a machine-learning based model that predicts user gestures, allowing the device to perform actions based on these predictions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple sensors are used for gesture recognition, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The sensor system is segmented into distinct functional groups: biosignal sensors (accelerometers, gyroscopes, magnetometers) and non-biosignal sensors (optical sensors, capacitive sensors, ultrasonic sensors). Each sensor type captures specific gesture characteristics, allowing the system to achieve high measurement precision through specialized measurement while managing overall complexity through modular organization.
Solution Approach 2:
The machine-learning based gesture recognition system serves multiple functions simultaneously: it detects various gesture types (wrist movements, hand movements, finger movements), identifies different users through personalized models, and determines action triggers. This multi-functionality allows a single integrated system to achieve high precision across diverse gesture recognition tasks without proportionally increasing complexity.
2Adaptability or versatility
If machine-learning based models are used for gesture prediction, then adaptability is improved, but device complexity increases
Solution Approach 1:
Personalized gesture prediction models are created in advance through a training phase where the system learns user-specific gesture patterns and characteristics. Once trained, these models enable the system to accurately predict gestures for multiple users without requiring real-time complex analysis, thus improving adaptability while managing processing complexity through pre-computation.
Solution Approach 2:
The system incorporates feedback mechanisms where sensor data is continuously fed back to the machine-learning models to refine predictions and update user-specific gesture profiles. This feedback loop enables the system to adapt to variations in user behavior and improve accuracy over time, while the feedback is processed efficiently through optimized algorithms that balance adaptability with computational complexity.
Data Source
AI summary
A device implementing a system for machine-learning based gesture recognition includes at least one processor configured to, receive, sensor data for a first window of time and additional sensor data for a second window of time overlapping the first window of time. The sensor data and the additional sensor data are provided as inputs to a machine learning model, the machine learning model having been trained to output a predicted gesture, predicted gesture start time, and predicted gesture end time based on the sensor data. A predicted gesture is determined based on an output from the machine learning model, and to perform, in response to determining the predicted gesture, a predetermined action on the device.