MobileNet Drawing-Style Authentication for Device Unlocking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing unlocking methods for electronic devices, such as numeric passwords, pattern locks, fingerprint recognition, and facial recognition, are rigid and limit user experience and convenience.
Innovation Solution
A drawing-based unlocking method and device that utilizes a neural network recognition model, specifically MobileNet, to identify the unique drawing style of a user for identity verification and device unlocking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If traditional unlocking methods (numeric passwords, pattern locks, fingerprint recognition, facial recognition) are used, then device security is maintained, but user experience and convenience are limited
Solution Approach 1:
The patent changes the parameter of unlocking input from traditional fixed formats (numeric passwords, standard patterns) to freehand drawing inputs with variable characteristics (stroke pressure, speed, continuity, stylistic features). This allows users to draw any shape or pattern freely while the recognition model adapts to various drawing styles, thereby improving ease of operation and user experience without compromising security
2Ease of operation
If freehand drawing unlocking is implemented, then user experience and convenience are improved, but recognition accuracy may be affected by variations in drawing styles
Solution Approach 1:
The patent segments the drawing recognition task into multiple feature dimensions including stroke pressure, drawing speed, continuity, and stylistic characteristics. The recognition model analyzes each feature separately and integrates them to make a comprehensive authentication decision, thereby maintaining high recognition accuracy despite variations in overall drawing style
Solution Approach 2:
The system transforms the drawing input into multiple parametric features (stroke pressure, speed, continuity, stylistic parameters) that capture the essence of the user's drawing behavior. By recognizing patterns in these parameters rather than exact geometric shapes, the system achieves high accuracy while allowing flexible drawing styles
3Adaptability or versatility
If a recognition model is trained on multiple training drawings with different patterns, then the unlocking system becomes more adaptable to various drawing styles, but the training complexity and time required increase
Solution Approach 1:
The patent performs preliminary action by pre-training the recognition model during device setup or manufacturing, so that the model is already trained on diverse drawing patterns before actual use. This preliminary training phase, though time-consuming, is performed once rather than repeatedly, allowing fast and accurate recognition during subsequent unlocking operations without requiring users to wait for training
Data Source
AI summary
A drawing-based unlocking method is provided. The method is implemented by a device. The method includes training a recognition model. The method includes using the recognition model to identify whether a drawing style of a drawing belongs to an owner of the device when receiving the drawing that is input by a user. The method includes unlocking the device when the drawing style belongs to the owner. The method includes not unlocking the device when the drawing style does not belong to the owner.