Web Conference Audio Verification Using Transcript Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In web conferencing sessions, participants are often unaware of the audio quality received by other users, leading to potential disruptions due to deteriorated audio quality that can go undetected, especially when network connections and processing delays cause significant degradation.
Innovation Solution
An audio controller generates transcripts of audio data at both the sending and receiving client devices using speech recognition algorithms, compares these transcripts to assess audio quality, and provides notifications to the sender about deteriorated audio quality, identifying specific portions and affected devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If audio data is transmitted over network connections during web conference sessions, then real-time communication between multiple users is enabled, but audio quality deteriorates due to network issues and processing delays
Solution Approach 1:
The system performs preliminary actions by generating transcripts of audio data at both sending and receiving devices before comparing them. This advance preparation of transcripts enables quality assessment without delaying the real-time communication flow, resolving the contradiction between speed and reliability.
Solution Approach 2:
The system implements feedback by continuously monitoring audio quality through transcript comparison and providing notifications to users about quality issues. This feedback loop allows users to adjust their speech or network conditions in real-time, maintaining reliable communication despite network variations.
2Measurement precision
If transcript comparison is performed to verify audio quality, then audio quality issues can be detected, but system complexity increases due to additional processing requirements
Solution Approach 1:
The system creates text copies (transcripts) of the audio data at both sending and receiving devices. By comparing these text copies instead of analyzing the audio signals directly, the system achieves precise quality detection while using simpler text processing algorithms, thereby reducing overall system complexity.
3Loss of information
If speech recognition algorithms are applied to generate transcripts, then audio quality can be assessed through text comparison, but processing time and computational resources increase
Solution Approach 1:
The system generates transcripts as a preliminary step that occurs in parallel with audio transmission. By preparing the text representations in advance and comparing them concurrently with audio delivery, the system minimizes additional processing time while maintaining accurate audio information through the transcript comparison mechanism.
Data Source
AI summary
A method may for audio quality verification may include generating a first transcript of a first audio data captured at a first client device engaged in a web conference session with a second client device. A second transcript of a second audio data received and/or output at the second client device may be generated and analyzed relative to the first transcript. In the event an above-threshold difference is detected between the first transcript and the second transcript, a notification may be generated for display at the first client device. The notification may indicate that the audio quality at the second client device fails to satisfy a quality threshold. Moreover, in some cases, the notification may identify portions of the first audio data that was not output at the second client device with sufficient audio quality, thereby enabling a repeat of those portions of the first audio data.