Footwear Sensor Embedding for User Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods fail to objectively identify and differentiate between various activities practiced by individuals wearing footwear and to distinguish between multiple users, as they struggle to assess duration and intensity of activities accurately.
Innovation Solution
A method and kit that utilize sensors in footwear to generate raw sensor data streams, which are then processed using a bidirectional recurrent neural network to project data into an embedding space, allowing classification by a machine learning classifier to predict user or activity classes, with features like bidirectional LSTM networks and wireless connectivity for data processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional sensor data processing methods are used, then the system is simple to implement, but it fails to accurately identify multiple users and differentiate activities
Solution Approach 1:
The patent introduces an embedding space as an intermediary representation layer between raw sensor data and classification. The bidirectional LSTM network projects high-dimensional sensor data into a compressed embedding space, which then serves as input for the classifier. This intermediary representation enables accurate user and activity identification while managing computational complexity through dimensionality reduction.
2Productivity
If high-dimensional raw sensor data is processed directly, then all information is preserved, but processing time and computational resources increase significantly
Solution Approach 1:
The patent transforms the parameter representation of sensor data by changing from high-dimensional raw data to compressed embedding vectors. The bidirectional LSTM network learns optimal parameter transformations that preserve essential information for user and activity identification while reducing dimensionality. This parameter change enables efficient processing without significant information loss.
3Measurement precision
If multiple sensors are added to capture comprehensive movement data, then measurement completeness improves, but device complexity and cost increase
Solution Approach 1:
The patent segments the sensor data processing into distinct functional components: data acquisition from multiple sensors, temporal segmentation through sliding windows, feature extraction via bidirectional LSTM, and classification. This segmentation allows the system to effectively process multi-sensor data while managing complexity through modular architecture and specialized processing for each data segment.
Data Source
AI summary
The invention relates to the field involving the performance of activities. In particular, the invention relates to a method and kit for recognising a user of a footwear article or an activity performed by a user of a footwear article. One of the objectives of the invention is to facilitate the detection of a user of a footwear article or of an activity performed by a user of a footwear article. For this purpose, the inventors propose the use of raw sensor data from the footwear article, which data is supplied to a machine-learning-trained classifier. The inventors have discovered that the use of raw sensor data, without fusion, by a classifier allows surprisingly good results to be obtained in the recognition of a subject or of different activities performed by a subject. In particular, the invention uses the embedding technique to project the raw data in a data representation space that is suitable for the desired classification.