Context-Aided Identification Using Local Machine Learning Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing computer identification systems face challenges in accurately identifying individuals from larger populations due to increased computational burdens, complexity, and privacy concerns, especially when relying on global machine learning models hosted in the cloud.
Innovation Solution
The approach utilizes multiple local machine learning models hosted on smart devices, such as augmented reality glasses, which gather sensor data and context information to associate fingerprints with possible identities, updating a database with these associations to improve identification accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a global machine learning model is used to identify individuals from a larger population, then identification accuracy is improved, but computational resources and processing power requirements increase
Solution Approach 1:
The patent divides the identification system into multiple local machine learning models distributed across different devices (smartphones, computers, IoT devices) rather than using a single global model. Each local model handles identification for a specific device or user context, segmenting the computational workload and reducing the processing power required at any single point while maintaining overall identification accuracy.
2Measurement precision
If a global machine learning model is used to identify individuals from a larger population, then identification accuracy is improved, but device complexity increases
Solution Approach 1:
The patent implements local machine learning models tailored to specific devices and user contexts rather than deploying a complex global model everywhere. Each local model is optimized for its specific environment and device capabilities, reducing overall system complexity while maintaining identification accuracy through context-specific adaptations.
3Measurement precision
If a global machine learning model is hosted in the cloud, then identification accuracy is improved, but latency increases due to internet connectivity requirements
Solution Approach 1:
The patent performs identification using local machine learning models before cloud synchronization is needed. By making identification decisions locally in real-time and only synchronizing results to the cloud afterward, the system eliminates the latency associated with cloud-based processing while maintaining accurate identification through locally trained models.
4Measurement precision
If a global machine learning model is used, then identification accuracy for larger populations is improved, but storage requirements increase
Solution Approach 1:
The patent segments the storage burden by distributing model parameters and data across multiple local devices rather than centralizing everything in the cloud. Each device stores only the local model and relevant local data, significantly reducing individual storage requirements while collectively maintaining the capability to identify individuals from large populations through the distributed network.
Data Source
AI summary
Smart devices can be configured to collect and share various forms of context data about where a user is located (e.g., location), what a user will be doing (e.g., schedule), and what a user is currently doing (e.g., activity). This context data may be combined with fingerprint data (e.g., biometrics) to help identify the fingerprint data. For example, a location of a user may help associated speech detected at that location with the user. These associations may be stored in a mapping database that can be updated over time to reduce ambiguities in identification. The mappings in the database may be used to train a machine learning model to recognize fingerprints as identities, which may be useful in applications, such as speaker identification.