Captioning Relay Queueing for Accurate Low-Cost Call Transcription
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional telecommunication systems for hearing impaired individuals face challenges in providing accurate, real-time voice-to-text captioning due to limitations in automated transcription software, high CA turnover rates, and the need for trained human operators, leading to delays and increased costs.
Innovation Solution
A semi-automated relay system that combines automated voice-to-text transcription with human assistance, utilizing remote ASR servers and local device processors to enhance accuracy and reduce reliance on human operators, with features like voice modeling and error correction to improve captioning efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If automated transcription software is used, then operational costs are reduced, but transcription accuracy deteriorates
Solution Approach 1:
The system segments the transcription task into two parts: automated transcription by ASR software for cost efficiency, and manual review/correction by human operators for accuracy. This division allows the system to leverage the strengths of both automated and human processes while mitigating their respective weaknesses.
Solution Approach 2:
The system introduces an intermediary review process where human operators verify and correct automated transcriptions. This intermediary layer acts as a bridge between automated cost-saving measures and the need for accurate transcription, allowing both goals to be achieved simultaneously.
2Measurement precision
If human operators are used for transcription, then transcription accuracy is improved, but operational costs increase
Solution Approach 1:
Instead of using human operators for complete transcription (excessive action), the system uses them only for reviewing and correcting automated transcriptions (partial action). This reduces operational costs while maintaining accuracy, as human involvement is limited to correcting errors rather than transcribing from scratch.
Solution Approach 2:
The automated transcription system serves itself by generating initial transcriptions that are then reviewed by human operators. This self-service approach reduces the burden on human operators, allowing them to focus only on correction rather than complete transcription, thereby reducing operational costs.
3Measurement precision
If more human operators are hired, then transcription accuracy is improved, but device complexity increases
Solution Approach 1:
The system extracts the transcription function from human operators and assigns it to automated ASR software. This reduces system complexity by removing the need for multiple human operators while maintaining accuracy through the automated review process. The complex task of full transcription is taken out of the human domain and handled by software.
4Productivity
If automated transcription is used, then productivity is improved, but transcription quality deteriorates
Solution Approach 1:
The system implements a feedback mechanism where automated transcriptions are reviewed and corrected by human operators. The corrected transcriptions then serve as training data to improve the ASR software's accuracy over time. This feedback loop maintains high productivity while progressively improving transcription quality.
Solution Approach 2:
The automated ASR software performs preliminary transcription before human review. This preliminary action captures the bulk of the transcription work at high speed, while subsequent human review ensures quality. The preliminary automated transcription maintains productivity while the follow-up review ensures accuracy.
Data Source
AI summary
A captioning relay for captioning hearing user (HU) voice signals comprising a plurality of separate captioning resources and a captioning administrator module that receives HU voice signal segments corresponding to a plurality of separate ongoing calls between HUs and AUs and provides the voice signal segments in a first in, first out order to the captioning resources, the administrator module providing each voice signal segment from each call to any one of the captioning resources to be captioned without regard to which captioning resource captioned prior voice signal segments generated during the call and, the administrator module further receiving caption segments back from the captioning resources and providing those captioning segments to AU devices associated with the calls that generated corresponding HU voice signal segments, and wherein the number of captioning resources is less than the number of ongoing calls.