Force Touch Identification Using Dynamic Deep Learning Model Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional electronic devices require a separate force touch sensor, which increases costs and hinders product downsizing, and deep learning models for identifying force touches suffer from reduced accuracy due to high computation rates.
Innovation Solution
An electronic device that selects an appropriate deep learning model from a plurality of models based on a history of previous force touch determinations to identify whether a touch input is a long touch or a force touch, using either a high-computation model for accuracy or a low-computation model for responsiveness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a force touch sensor is added to identify force touches, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent extracts the force touch identification function from the dedicated force touch sensor and relocates it to the existing touchscreen's processing system. By using deep learning models that process touch pixel data from the touchscreen itself, the system eliminates the need for separate force touch sensors while maintaining identification capability.
Solution Approach 2:
The touchscreen's processing system is enhanced to perform multiple functions: it not only detects touch location but also identifies force touch characteristics through deep learning models. This multi-functional approach allows the existing touchscreen hardware to handle both basic touch detection and force touch identification without additional dedicated sensors.
2Speed
If a high-computation rate deep learning model is used for rapid force touch response, then speed is improved, but measurement precision deteriorates
Solution Approach 1:
The system dynamically selects between different deep learning models based on the specific requirements of each force touch identification task. The processor can switch between high-computation models for accuracy-critical scenarios and lower-computation models for speed-critical scenarios, making the system adaptable to varying performance requirements.
Solution Approach 2:
The system changes the computation parameters by selecting different deep learning models with varying computational complexities. This allows optimization of either speed or precision depending on the operational context, such as selecting a higher-computation model when precision is prioritized or a lower-computation model when response speed is critical.
Data Source
AI summary
According to various embodiments, an electronic device includes a memory storing deep learning models for determining a force touch, a touchscreen, and a processor configured to identify a touch input of a user through the touchscreen, receive touch pixel data for frames having a time difference based on the touch input, and identify whether the touch input is a force touch based on the touch pixel data. The processor is configured to identify whether the touch input is the force touch using a first determination model among the deep learning models in response to identifying that the touch input is reinputted a designated first number of times or more within a designated time, and otherwise, identify whether the touch input is the force touch using a determination model having a lower computation load than the first determination model among the deep learning models.