Multi-touch Gesture Recognition for Graphical Keyboards
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Graphical keyboards on computing devices suffer from limitations in accuracy, speed, and adaptability, making text input tedious, slow, and error-prone, especially for longer words, and existing methods require discrete gestures for each key.
Innovation Solution
A method that outputs a graphical keyboard and detects multi-touch gestures comprising sub-gestures traversing groups of keys, determining candidate words by analyzing the points traversed by each sub-gesture and associating probabilities with keys, allowing incremental and real-time prediction of text entries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If discrete gestures are used for each key, then the method is simple to implement, but the text input speed is slow and tedious
Solution Approach 1:
The patent segments the gesture recognition process into sub-gestures, where each sub-gesture corresponds to a sequence of key traversals. This allows the system to process complex multi-key gestures as manageable units, improving recognition speed and accuracy while maintaining implementability through structured processing stages.
Solution Approach 2:
The patent performs preliminary actions by pre-defining gesture models and probability thresholds before actual text input occurs. The system pre-processes gesture patterns and associates them with candidate words, enabling rapid recognition during actual typing without complex real-time calculations.
2Measurement precision
If traditional graphical keyboards are used, then the interface is simple, but accuracy and adaptability are limited
Solution Approach 1:
The patent introduces dynamic adaptability by adjusting probability thresholds and gesture model parameters based on user behavior patterns. The system dynamically refines its recognition accuracy over time through adaptive learning, allowing it to accommodate varying user styles and improve precision without requiring a completely complex static interface.
Solution Approach 2:
The patent implements feedback mechanisms where the system provides real-time candidate word predictions based on detected sub-gestures. This feedback loop allows the system to continuously refine its accuracy by comparing predicted words with actual user input, gradually improving recognition precision while maintaining a relatively simple visual keyboard interface.
3Productivity
If multi-touch gestures with multiple sub-gestures are detected, then the rate of text input is enhanced, but the processing time for post-gesture analysis increases
Solution Approach 1:
The patent performs preliminary processing by pre-calculating gesture models and probability distributions before the actual gesture occurs. The system pre-establishes the relationship between sub-gesture patterns and candidate words, enabling rapid post-gesture processing without time-consuming real-time analysis.
Solution Approach 2:
The patent maintains continuous processing during the gesture itself rather than waiting for the gesture to complete. The system continuously evaluates sub-gestures and updates candidate word predictions in real-time during the gesture execution, eliminating the need for separate post-gesture processing stages and maintaining continuous useful action throughout the interaction.
Data Source
AI summary
In one example, a method comprises outputting, by a computing device and for display at an output device, a graphical keyboard comprising a plurality of keys, and receiving, by the computing device, an indication of a multi-touch gesture detected at a presence-sensitive display, the multi-touch gesture comprising a first sub-gesture that traverses a first group of keys of the plurality of keys and a second sub-gesture that traverses a second group of keys of the plurality of keys. This example method further comprises determining, in response to detecting the first sub-gesture and the second sub-gesture, a candidate word based at least in part on the first and second groups of keys, and outputting, by the computing device and for display at the output device, the candidate word.