Wearable EMG Gesture Identification via Decision Tree Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Wearable electromyography (EMG) devices face challenges in performing real-time automated gesture identification due to the computational intensity of pattern recognition and machine learning algorithms, which requires significant processing power and resources, leading to battery life issues and slow processing speeds.
Innovation Solution
A wearable EMG device with a processor that performs decision tree analysis on EMG signal features to determine probability scores for gesture identification, using a series of evaluations based on previous outcomes, and combines signals from multiple time windows to enhance accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If computationally intensive pattern recognition and machine learning algorithms are used for gesture identification, then gesture identification accuracy is improved, but processing speed decreases and battery life is reduced
Solution Approach 1:
The gesture identification process is segmented into multiple evaluation stages. The system performs a series of evaluations of EMG signal features in sequence, where each evaluation builds upon the outcome of the previous one. This segmentation allows the system to achieve high accuracy through comprehensive feature analysis while maintaining manageable processing speed by evaluating features progressively rather than simultaneously.
Solution Approach 2:
The system performs preliminary feature extraction and evaluation before final gesture identification. By determining a set of features from EMG signals and performing evaluations on these features in advance, the system prepares the necessary information for accurate gesture identification without requiring computationally intensive real-time processing during the actual gesture recognition.
2Measurement precision
If computationally intensive algorithms are used for gesture identification, then gesture identification accuracy is improved, but energy consumption increases
Solution Approach 1:
The energy-intensive gesture identification task is divided into sequential evaluation stages. Each evaluation processes specific EMG signal features and builds upon previous results, allowing the system to achieve accurate gesture identification through cumulative, manageable computational steps rather than a single intensive processing burst, thereby extending battery life.
Solution Approach 2:
The system performs evaluations of EMG signal features in periodic cycles, where each cycle consists of determining features, performing evaluations, and updating probability scores. This periodic processing allows the system to maintain gesture identification accuracy while managing energy consumption through structured, repeatable processing cycles that can be optimized for power efficiency.
3Productivity
If complex processor infrastructure is used to support fast processing, then processing speed is improved, but device complexity increases
Solution Approach 1:
The processing function is segmented into discrete evaluation steps that can be performed by a simple processor. Instead of requiring a complex processor to handle all processing tasks simultaneously, the system breaks down the processing into sequential evaluations of EMG signal features, allowing a simple processor to achieve fast processing through efficient step-by-step execution.
Solution Approach 2:
The system extracts and evaluates specific features from EMG signals separately from the main processing task. By determining a set of features and performing evaluations on these extracted features in advance, the system reduces the computational burden on the processor during gesture identification, enabling fast processing with minimal processor infrastructure.
Data Source
AI summary
Systems, articles, and methods perform gesture identification with limited computational resources. A wearable electromyography (“EMG”) device includes multiple EMG sensors, an on-board processor, and a non-transitory processor-readable memory that stores data and/or processor-executable instructions for performing gesture identification. The wearable EMG device detects and determines features of signals when a user performs a physical gesture, and processes the features by performing a decision tree analysis. The decision tree analysis invokes a decision tree stored in the memory, where storing and executing the decision tree may be managed by limited computational resources. The outcome of the decision tree analysis is a probability vector that assigns a respective probability score to each gesture in a gesture library. The accuracy of the gesture identification may be enhanced by performing multiple iterations of the decision tree analysis across multiple time windows of the EMG signal data and combining the resulting probability vectors.