Wearable Gesture Control for Multi-Device Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing remote control systems for multiple electronic devices in a user environment face challenges such as incompatibility, need for line-of-sight, and user discomfort with voice commands, leading to high user friction and inconvenience.
Innovation Solution
A wearable device equipped with IMU and PPG sensors generates gesture signals processed by a convolutional neural network (CNN) to identify and control multiple electronic devices within a vicinity, allowing universal gesture-based control without line-of-sight requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a gesture-based remote control is configured to control all devices in the room, then the user can control multiple electronic devices, but the system becomes complex in identifying which device the user intends to control
Solution Approach 1:
The patent introduces an intermediary system consisting of wearable sensors and machine learning models that mediate between the user's gesture and the target device. The wearable device captures gesture data, and the ML model processes this data to identify both the intended device and the desired control action, thereby resolving the complexity of direct multi-device control selection
Solution Approach 2:
The system changes the parameter of device identification from traditional methods (remote control codes, voice commands) to gesture-based identification. By using wearable sensors to capture motion parameters and applying ML models, the system transforms how device selection is made, simplifying the user's interaction while maintaining the ability to control multiple devices
2Ease of operation
If traditional remote controls are used for each device, then device control is straightforward, but the user experiences high user friction due to incompatibility and need for multiple remotes
Solution Approach 1:
The patent implements a universal gesture-based control system that can control multiple different types of electronic devices with a single wearable device. The machine learning model is trained to recognize gestures intended for different devices and translate them into appropriate control commands, eliminating the need for multiple device-specific remotes and reducing user friction
Solution Approach 2:
The system replaces traditional mechanical remote controls with a gesture-based control mechanism. Instead of physically pointing a remote at a device or pressing buttons, the user performs natural gestures captured by wearable sensors, and the ML model translates these gestures into device control commands, simplifying the interaction model
3Ease of operation
If voice commands are used for device control, then hands-free operation is achieved, but users experience discomfort and privacy concerns
Solution Approach 1:
The patent substitutes voice-based acoustic control with gesture-based mechanical motion control. Instead of using the user's voice, the system uses wearable sensors to capture hand and body gestures, translating these mechanical movements into device control commands. This provides hands-free operation while avoiding the privacy and comfort issues associated with voice commands
Data Source
AI summary
Techniques of controlling electronic devices using gestures use a wearable device on a user which translates, via a model, user movements into signals that both identify an electronic device to be controlled and a specific action to take with regard to that electronic device. The wearable device includes an inertial measurement unit (IMU) sensor and a photoplethysmography (PPG) sensor and measure six degrees of freedom (6DOF). The model is a convolutional neural network (CNN) that takes x, y, and z-acceleration signals generated by the IMU and PPG and places each acceleration component generated from each sensor in a separate channel. The CNN takes the input from each channel and generates a respective, separate model for each channel. The output at each of the stacked layers are combined in a fully connected layer to produce CNN output identifying an electronic device and a control for the electronic device.