Behavior-Based Authentication Switching for Gait and No-Lock Access
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing information processing apparatuses with personal authentication functions, such as fingerprint or face authentication, require frequent user interaction for unlocking, leading to inconvenience, especially in portable devices like smartphones, where unlocking operations are performed multiple times daily.
Innovation Solution
An information processing apparatus that estimates user behavior using a behavior estimation library and switches authentication methods between gait authentication, other authentication methods, and no lock based on the estimated behavior, using a neural network to generate a gait model and a behavior rule database to determine appropriate authentication strategies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional authentication methods (fingerprint, face authentication) are used, then security protection is achieved, but user convenience deteriorates due to frequent authentication operations required
Solution Approach 1:
The patent implements dynamic switching between multiple authentication methods based on real-time behavior detection. The system transitions from static authentication (requiring user initiation) to dynamic authentication (automatically detecting gait patterns), adapting the authentication approach according to the user's current state and context.
Solution Approach 2:
The system enables self-service authentication by automatically detecting and analyzing the user's gait pattern without requiring explicit user initiation. The authentication process occurs autonomously in the background, eliminating the need for users to manually trigger authentication operations while maintaining security.
2Reliability
If authentication operations are performed frequently to maintain security, then security protection is improved, but time consumption increases
Solution Approach 1:
The system performs preliminary authentication by continuously monitoring and analyzing gait patterns in the background before explicit authentication is needed. This preliminary detection allows the system to pre-validate the user's identity, eliminating the need for time-consuming authentication operations at the moment of access.
Solution Approach 2:
The authentication system operates continuously in the background through ongoing gait analysis, rather than intermittently triggering discrete authentication events. This continuous monitoring maintains security validation without requiring periodic user interruptions, thereby reducing time loss while preserving protection.
3Ease of operation
If gait authentication is implemented to reduce authentication operations, then user convenience is improved, but system complexity increases due to behavior estimation library and neural network requirements
Solution Approach 1:
The patent leverages existing multi-functional components (sensors, processors, machine learning frameworks) to implement gait authentication. These components serve multiple purposes including device operation monitoring, user interaction detection, and authentication validation, thereby avoiding the need for dedicated specialized hardware and reducing overall system complexity.
4Adaptability or versatility
If behavior-based authentication switching is implemented, then authentication adaptability is improved, but computational requirements increase due to neural network processing
Solution Approach 1:
The system implements partial authentication processing by using lightweight gait feature extraction and comparison rather than full neural network inference for every authentication check. The complex neural network model is trained offline and deployed as a simplified decision framework, performing only essential classification tasks during runtime to minimize energy consumption while maintaining adaptability.
Data Source
AI summary
An information processing apparatus includes a behavior estimation library and an authentication method switching unit. The behavior estimation library estimates a behavior of a user. The authentication method switching unit switches, based on an estimation result of the behavior, an authentication method of personal authentication among a gait authentication method, an authentication method other than the gait authentication method, and no lock.


