Audio Detection System for Video Conferencing Speaker Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automated audio detection systems in video conferencing often inaccurately identify participants as speakers due to non-speech audio, such as background noise, leading to unreliable speaker identification and user experience issues.
Innovation Solution
Implementing a method that combines audio detection and speech recognition to differentiate between speech and non-speech audio, providing accurate identification of speakers and non-speakers, and enabling remedial actions like notifications to reduce non-speech audio sources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional audio detection is used to identify speakers, then the system can detect audio presence, but it cannot accurately distinguish between speech and non-speech audio
Solution Approach 1:
The patent combines audio detection and speech recognition into a unified detection system. The audio detection component identifies audio presence while the speech recognition component determines whether the audio is actual speech. By merging these two components and applying a logical function to their outputs, the system achieves accurate speaker identification that distinguishes speech from non-speech audio sources.
2Productivity
If audio detection threshold is lowered to detect more speakers, then more speakers are identified, but false positives increase due to non-speech audio
Solution Approach 1:
The system uses speech recognition output as feedback to validate audio detection results. When audio detection identifies a potential speaker, the speech recognition component provides feedback to confirm whether the audio is actual speech. This feedback mechanism allows the system to maintain high detection coverage while filtering out false positives from non-speech audio sources.
3Measurement precision
If speech recognition is added to audio detection, then speech accuracy improves, but system complexity and processing time increase
Solution Approach 1:
The patent segments the detection system into two independent but coordinated components: audio detection and speech recognition. Each component operates semi-independently, with the audio detection component handling audio presence detection and the speech recognition component handling speech verification. This segmentation allows the system to achieve high speech detection accuracy while managing complexity through modular architecture.
4Measurement precision
If speech recognition is applied to all audio streams, then speaker accuracy improves, but processing time and computational resources increase
Solution Approach 1:
The system performs preliminary audio detection on all audio streams to identify potential speakers before applying speech recognition. This preliminary action filters out streams that clearly do not contain speech, allowing speech recognition to be applied selectively only to streams that pass the audio detection threshold. This approach maintains high speaker identification precision while reducing overall processing time and computational resource consumption.
Data Source
AI summary
A computer system executing a computer audio application such as video conferencing applies audio detection and speech recognition to an input audio stream to generate respective audio detection and speech recognition signals. A function is applied to the audio detection and speech recognition signals to generate a non-speech audio detection signal identifying presence of non-speech audio in the input audio stream when the audio detection signal is asserted and the speech recognition signal is not asserted. A control or indication action is performed in the computer system based on assertion of the non-speech audio detection signal.