Gaze-Based Text Input Using Eye Tracking and Decoding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing gaze-based text input methods are slow and inaccurate, and speech recognition software may not be convenient or reliable for users without hand control or speech disorders.

Innovation Solution

A system and method for entering text into a computing device using gaze input, which includes an eye tracking device to monitor and process gaze data, a decoding module to hypothesize text based on gaze patterns, and error correction interfaces to improve accuracy, allowing users to input text without physical contact.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If speech recognition software is used for text input, then text input can be provided without physical contact, but it may be too slow, require extensive training, or not be available to users with speech disorders

Engineering Contradiction:
Improvetext input accessibilityVSAvoidtext input speed
Core Design Contradiction:
Ease of operationVSProductivity

Solution Approach 1:

The patent replaces speech recognition (acoustic field) with gaze-based input (optical field). The eye tracking device detects gaze location through optical means, and the system translates gaze patterns into text input without requiring speech processing. This substitution eliminates the limitations of speech recognition while maintaining contactless operation.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Ease of operation

If basic gaze-based text input methods are used, then physical contact is not required, but the input speed and accuracy suffer

Engineering Contradiction:
Improvecontactless text inputVSAvoidtext input accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The system implements feedback through error correction interfaces that present multiple likely terms based on detected gaze patterns. The user can review and correct the interpreted text, and the system learns from corrections to improve future accuracy. This feedback loop significantly enhances text input precision while maintaining contactless operation.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The system performs preliminary processing by analyzing gaze patterns and generating multiple likely term hypotheses before final text confirmation. This preliminary action allows the system to prepare accurate text suggestions in advance, improving both speed and accuracy of the final text input.

Inventive Principle:
Principle #10Preliminary action

3Ease of operation

If basic gaze-based text input methods are used, then physical contact is not required, but the input process becomes slow

Engineering Contradiction:
Improvecontactless text inputVSAvoidtext input time
Core Design Contradiction:
Ease of operationVSLoss of time

Solution Approach 1:

The system performs preliminary analysis of gaze patterns and generates multiple likely term hypotheses before final confirmation is needed. By preparing multiple probable text interpretations in advance based on the gaze sequence, the system reduces the time required for text input validation and correction, thereby reducing overall text input time.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The error correction interface provides immediate feedback with multiple likely terms, allowing users to quickly confirm or correct text without lengthy processing delays. This feedback mechanism accelerates the text input process by presenting ready-made corrections rather than requiring complete re-analysis.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3189398B1Gaze based text input systems and methods
Publication Date: 2020.03.18 TOBII TECH AB
  • EP3189398B1 patent drawingFigure 1
  • EP3189398B1 patent drawingFigure 2
  • EP3189398B1 patent drawingFigure 3

AI summary

According to the invention, a method for entering text into a computing device using gaze input from a user is disclosed. The method may include causing a display device to display a visual representation of a plurality of letters. The method may also include receiving gaze information identifying a movement of the user's gaze on the visual representation. The method may further include recording an observation sequence of one or more observation events that occur during the movement of the user's gaze on the visual representation. The method may additionally include providing the observation sequence to a decoder module. The decoder module may determine at least one word from the observation sequence representing an estimate of an intended text of the user.