Speaker identification-based echo detection and solution
US12069113B2 · kind B2 · utility
Assignee
Inventors
Key dates
| Filing date | Oct 11, 2022 |
| Grant date | Aug 20, 2024 |
| Priority date | — |
| Expiry date | Oct 11, 2042 |
Classification
- Technology area (CPC H)Electricity
- CPC primaryH04M9/082
- WIPO fieldDigital communication
- WIPO sectorElectrical engineering
Abstract
A video conference call control system is provided with an audio mixer to receive an audio speech signal at one or more network devices connected to the video conference call; to generate first and second talker identification values from the audio speech signal; to compare the first and second talker identification values during a timing comparison window to determine if the first and second talker identification values are identical and received simultaneously or substantially contemporaneously; and to selectively mute one or more audio devices at the network device(s) if the first and second talker identification values are identical and received simultaneously or substantially contemporaneously so that microphone input signals to the audio mixer are muted to prevent “playback” echo effects and so that microphone input and audio loudspeaker output signals to the audio mixer are muted to prevent “double join” echo effects.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.