Speech Recognition Text Editing Queue

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing dictation tools require users to frequently switch between recording and editing modes, leading to inefficiencies and user frustration, particularly in applications like creating medical reports where multiple text manipulations are needed.

Innovation Solution

A method and system that allows users to edit text simultaneously while recording speech, by queuing editing actions in the audio queue, enabling instant visual changes without interrupting the recording process, and processing these changes once the initial speech is converted to text.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If users frequently switch between recording mode and editing mode to manipulate text, then text editing capability is improved, but user-friendliness deteriorates and the number of button clicks increases

Engineering Contradiction:
Improvetext editing capabilityVSAvoiduser-friendliness
Core Design Contradiction:
Adaptability or versatilityVSEase of operation

Solution Approach 1:

The patent merges the recording mode and editing mode into a single integrated mode. Users can perform text manipulations (selection, deletion, over-diction) while speech recording is actively running, eliminating the need to switch between separate recording and editing modes. This combines two previously separate functions into one unified operational state.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The system queues text manipulation requests in advance and processes them as speech segments become available. When a user performs a text manipulation during recording, the request is queued and executed preliminarily once the corresponding speech segment is recognized, allowing the user interface to respond immediately while the actual processing occurs in the background.

Inventive Principle:
Principle #10Preliminary action

2Manufacturing precision

If users switch between recording mode and editing mode for each text manipulation, then text accuracy is improved, but productivity deteriorates due to repeated mode switching

Engineering Contradiction:
Improvetext accuracyVSAvoidreport creation speed
Core Design Contradiction:
Manufacturing precisionVSProductivity

Solution Approach 1:

The patent enables continuous speech recording while text manipulations are performed and processed in real-time. The speech recognition engine continues processing speech segments uninterrupted, and text manipulations are applied as they occur without requiring pauses or mode switches, maintaining continuous productive action throughout the dictation process.

Inventive Principle:
Principle #20Continuity of useful action

3Ease of operation

If text manipulations are processed immediately during speech recording, then user-friendliness is improved, but speech recognition accuracy may deteriorate due to simultaneous processing

Engineering Contradiction:
Improveuser-friendlinessVSAvoidspeech recognition accuracy
Core Design Contradiction:
Ease of operationVSManufacturing precision

Solution Approach 1:

Text manipulation requests are queued in advance and processed preliminarily once the corresponding speech segments are fully recognized. The system prepares the manipulation operations beforehand but delays their actual execution until the speech recognition is complete, ensuring accuracy is not compromised while maintaining responsive user interaction.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent segments the processing into distinct phases: speech recording and recognition occurs in one thread, while text manipulations are queued and processed in another thread once speech segments are complete. This segmentation allows independent processing of speech recognition and text manipulation without interference between the two operations.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3061001B1Speech recognition method and system with simultaneous text editing
Publication Date: 2018.08.08 AGFA HEALTHCARE NV
  • EP3061001B1 patent drawingFigure 1~2
  • EP3061001B1 patent drawingFigure 3A~3G

AI summary

In order to generate text from audio input, speech from a user is stored in an audio queue (201), stored speech is transformed into text through speech recognition, and the text is displayed to the user. A text editing event (106, 107) inputted by the user is also stored in the audio queue (20 ), and changes resulting from the text editing event (201) are instantly displaying to the user. When all speech queued prior to the text editing event (107) in the audio queue (20 ) is transformed into text, speech recognition is halted and the text editing event is processed (111) while additional speech from the user is stored in the audio queue (201). As soon as the text editing event has been processed, speech recognition is resumed.