Auto-Encoder Model for Human Movement Validity Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing human body movement monitoring methods are prone to false alarms and inability to accurately detect fraudulent activities due to their reliance on single-dimensional information, such as step count, which is insufficient for determining the validity of movements.
Innovation Solution
A neural network-based method using an auto-encoder detection model that transforms n-dimensional movement data into feature text, comprising direction, distance, and step count, to determine movement validity by comparing initial and predicted encodings against a preset threshold, trained with datasets including GPS, sensor, and health information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If step counting by smart wearables is used to monitor human body movement, then movement records can be generated, but the system becomes susceptible to spoofing and produces false alarms
Solution Approach 1:
The patent transitions from single-dimensional step counting to multi-dimensional movement analysis by incorporating GPS coordinates, accelerometer data, gyroscope data, and other sensor information. This dimensional expansion enables more reliable fraud detection by examining movement from multiple angles simultaneously.
Solution Approach 2:
The patent combines multiple data sources and sensor types into a composite detection system. By fusing GPS, accelerometer, gyroscope, and other sensor data, the system creates a composite movement profile that is resistant to spoofing and enables accurate validity detection.
2Ease of operation
If simple step counting methods are used, then the system is easy to operate, but it cannot accurately distinguish valid movements from fraudulent activities
Solution Approach 1:
The system automatically collects, processes, and analyzes multi-dimensional movement data without requiring user intervention. The neural network model performs automatic fraud detection by comparing actual movement patterns against expected patterns, eliminating the need for manual verification while maintaining high precision.
Solution Approach 2:
The patent replaces manual fraud detection methods with an automated neural network-based system. The machine learning model automatically analyzes sensor data and GPS information to detect fraudulent activities, substituting complex manual analysis with intelligent automated processing.
3Device complexity
If traditional detection methods are used, then the system complexity is low, but the false alarm rate increases and detection accuracy decreases
Solution Approach 1:
The patent implements a dynamic detection system that adapts to different movement scenarios and user behaviors. The neural network model learns from historical data and adjusts its detection thresholds and parameters dynamically, enabling reliable fraud detection across varying conditions while managing system complexity through adaptive algorithms.
Data Source
AI summary
The present disclosure relates to a neural network-based method for detecting validity of a human movement, comprising, obtaining a user movement dataset; extracting a feature text based on the user movement dataset; wherein the feature text comprises a plurality of input sequences, and each of the plurality of input sequences comprises at least a direction, a distance, and a step count; transforming the feature text into a first initial encoding according to a preset coding rule; and inputting the first initial encoding into an auto-encoder neural network-based detection model to determine whether the human body movement corresponding to the user movement dataset is valid.