Mobility Device Emergency Detection Using Joint Tracking and Voice Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies fail to automatically determine and respond to violent situations in mobility devices, such as buses, lacking a technical means to proactively manage emergencies.
Innovation Solution
A method and system utilizing deep learning engines to analyze video and audio signals from multiple sensors to detect violent situations and emotional states, integrating passenger identification, joint position tracking, and emotional state analysis to trigger emergency measures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If existing black box technology is used to film assailants, then evidence collection is improved, but automatic emergency detection and response capability is insufficient
Solution Approach 1:
The system performs preliminary analysis of video and audio data continuously before emergencies occur. Deep learning models pre-process sensor inputs to detect potential violent situations and emotional states, enabling proactive emergency detection rather than reactive recording after the fact.
Solution Approach 2:
The mobility device autonomously analyzes its own sensor data through integrated deep learning engines without external intervention. The system self-determines emergency conditions by processing video, audio, and sensor inputs, then automatically executes response measures without requiring external monitoring or manual assessment.
2Measurement precision
If multiple sensors and deep learning engines are integrated for automatic emergency determination, then emergency detection accuracy is improved, but device complexity increases
Solution Approach 1:
The system divides the emergency detection function into separate specialized modules: image sensor processing for violent situation detection, microphone sensor processing for emotional state analysis, and vibration sensor processing for physical disturbance detection. Each module uses dedicated deep learning engines optimized for specific tasks, making the complex system modular and manageable.
Solution Approach 2:
Multiple sensor inputs (video, audio, vibration) are merged and processed simultaneously through coordinated deep learning engines that share computational resources and decision-making logic. The integration of these sensors creates a unified emergency detection capability that achieves higher accuracy without proportionally increasing complexity.
Data Source
AI summary
The present disclosure relates to a method of automatically managing an emergency in a mobility device and system for the same. The method includes distinguishing passengers through an image obtained by an image sensor and determining a violent situation based on per-passenger joint position tracking, determining emotional states of one or more passengers through voice obtained by a sound sensor, based on determining the emergency requiring an emergency measure based on the determination on the violent situation and the determination on the emotional states, storing the image of the image sensor and the voice of the sound sensor, and transmitting a signal for the emergency measure to an emergency measure server.