Teleconference Co-Located Device Detection for Audio Feedback Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Participant computing devices in teleconferencing sessions struggle to identify co-located devices, leading to audio feedback loops like 'howling' due to the obscuring of audio sources in mixed audio data, which complicates corrective actions and requires substantial computing resources.
Innovation Solution
Participant computing devices receive mixed audio data, compare it with their own captured audio, identify 'featured' devices contributing most actively to the mixed audio, and obtain co-location information to determine the co-located device, allowing for targeted corrective actions to reduce or eliminate audio feedback.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If mixed audio data from multiple devices is used in teleconferencing, then audio coverage and participation are improved, but audio feedback loops and howling occur due to obscured audio sources
Solution Approach 1:
The system continuously monitors mixed audio data for feedback signals and uses this feedback information to identify co-located devices and adjust audio processing in real-time, thereby resolving the contradiction between comprehensive audio coverage and the prevention of feedback loops
Solution Approach 2:
The patent introduces an intermediary processing layer that analyzes mixed audio to identify co-located devices and mediates between the audio sources and participants, enabling selective audio routing that prevents feedback while maintaining coverage
2Reliability
If co-location detection is performed to identify feedback sources, then corrective actions can be taken, but computing resources are consumed
Solution Approach 1:
The system extracts only the essential information from mixed audio data needed for co-location detection, such as timing information and audio fingerprints, rather than processing the entire audio stream, thereby reducing computing resource consumption while maintaining reliable feedback detection
Solution Approach 2:
The audio processing is segmented into discrete analysis phases where the system processes audio data in chunks and only performs full analysis when feedback patterns are detected, optimizing the balance between reliability and computational efficiency
3Measurement precision
If all participant devices are monitored for co-location, then accurate identification is achieved, but system complexity increases
Solution Approach 1:
Instead of continuously monitoring all devices, the system applies partial action by only performing comprehensive co-location analysis when feedback patterns are detected or when specific devices are identified as potential sources, reducing system complexity while maintaining identification accuracy
Data Source
AI summary
A participant computing device connected to a teleconferencing session receives mixed audio data from a teleconference computing system orchestrating the session. The mixed audio data comprises a mix of audio captured by other devices connected to the session. The mixed audio and audio captured by the other devices prior to receipt of the mixed audio data are compared to determine that the participant computing device is co-located with at least one of the other devices. Information indicating that some of the other devices are featured devices that most actively contribute to the mixed audio is obtained. Co-location information for each of the featured devices is received. The co-location information is indicative of whether a featured device is co-located with the participant computing device. Based on the co-location information, a determination is made that one of the featured devices is co-located with the participant computing device.