Gesture Recognition with Noise Filtering to Reduce False Alarms
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Recognition systems experience false alarms due to noise influencing sensors, leading to reduced user confidence in trusting the systems, as they may execute commands when no motion is detected.
Innovation Solution
A gesture recognition method and system utilizing two neural networks: a first recognition neural network to classify gestures and a first noise neural network to determine if sensing signals are noise, thereby reducing false alarms by executing gestures only when signals are not noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If the recognition system executes commands based on sensing signals, then the system responds quickly to user gestures, but the false alarm rate increases due to noise influencing the sensor
Solution Approach 1:
The patent divides the recognition system into two separate neural networks: a first recognition neural network for gesture classification and a first noise neural network for noise detection. This segmentation allows each network to specialize in its specific function, enabling the system to maintain quick response through the recognition network while reducing false alarms through the dedicated noise detection network
Solution Approach 2:
The first noise neural network acts as an intermediary between the sensing signal and the gesture execution. It detects noise in the sensing signal and prevents false gesture recognition, thereby reducing false alarms while allowing genuine gestures to be processed by the first recognition neural network for quick execution
2Adaptability or versatility
If the recognition system processes all sensing signals, then comprehensive gesture recognition is achieved, but noise signals are misclassified as gestures
Solution Approach 1:
The patent segments the signal processing function into two distinct neural networks: one for gesture recognition and another for noise detection. This segmentation enables the system to maintain comprehensive gesture recognition capability while improving signal classification accuracy by having the noise neural network specifically identify and filter out noise signals
Solution Approach 2:
The first noise neural network provides feedback about the presence of noise in the sensing signal to the gesture recognition process. This feedback mechanism allows the system to adjust its classification decisions based on noise detection, thereby improving measurement precision without compromising gesture recognition versatility
Data Source
AI summary
A performing device of a gesture recognition system for reducing a false alarm rate executes a performing procedure of a gesture recognition method for reducing the false alarm rate. The gesture recognition system includes two neural networks. A first recognition neural network is used to classify a gesture event, and a first noise neural network is used to determine whether the sensing signal is the noise. Since the first noise neural network can determine whether the sensing signal is the noise, the gesture event may not be executed when the sensing signal is the noise. Therefore, the false alarm rate may be reduced.