Headset Wired Link and Auto-Labeling for Low Power
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Smart mobile accessories with motion sensors face high power consumption and data analysis delays due to wireless transmission, and manual data labeling is inefficient and prone to errors, affecting the accuracy of identification models.
Innovation Solution
A headset with a motion sensor, processor, and transceiver that senses user posture, determines correctness, and transmits results via a wired connection, reducing power consumption and enabling efficient label data generation using image and depth image processing to automate data labeling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If wireless transmission technology is used to transmit sensing data from smart mobile accessories to smartphones, then data transmission can be achieved, but power consumption increases significantly
Solution Approach 1:
The patent extracts the data transmission function from the wireless communication module and implements it through a wired connection instead. The sensing data is processed locally by the processor and transmitted through a physical cable connection to the external device, eliminating the need for wireless transmission and significantly reducing power consumption of the motion sensor accessory.
2Ease of manufacture
If manual data labeling is performed by persons according to their own determination standards, then label data can be generated, but labeling efficiency is low and determination errors occur
Solution Approach 1:
The system enables self-service labeling by automatically generating label data through image processing and recognition algorithms. The processor analyzes the captured images, identifies user postures, and automatically assigns labels to sensing data without requiring manual intervention, thereby dramatically improving labeling efficiency and consistency.
Solution Approach 2:
The patent replaces the manual mechanical labeling process with an automated computer vision system. Image capturing devices capture user postures, the processor analyzes these images using recognition algorithms, and automatically generates labels, substituting human judgment with automated image processing technology.
3Ease of manufacture
If manual data labeling is performed over a long process, then label data can be generated, but determination errors occur due to fatigue
Solution Approach 1:
The automated labeling system eliminates human fatigue by using computer vision algorithms to consistently and accurately label data throughout the entire labeling process. The processor maintains high determination accuracy regardless of processing duration, as the automated system does not experience fatigue like human annotators.
Data Source
AI summary
A method for generating label data is provided, and the method includes the following steps: sensing a posture of a user to generate a non-depth image; sensing the posture of the user to generate a depth image; sensing the posture of the user to generate motion data; and generating an identifying result of the posture according to the non-depth image and the depth image and labeling the motion data according to the identification result to generate label data.