Patent · US Active

Conversational placement of speakers at one endpoint

US10148913B2 · kind B2 · utility

7Cited by
3References
16Claims
0Family size

Assignee

Inventors

Key dates

Filing dateOct 23, 2017
Grant dateDec 4, 2018
Priority date
Expiry dateOct 23, 2037

Classification

  • Technology area (CPC G)Physics
  • CPC primaryG06T2207/30196
  • WIPO fieldAudio-visual technology
  • WIPO sectorElectrical engineering

Abstract

An automatic process for producing professional, directed, production crew quality, video for videoconferencing is described. Rule based logic is integrated into an automatic process for producing director quality video for videoconferencing. The automatic process uses sensor data to process video streams for video conferencing. A method and system for automatically processing sensor data on room activity into general room analytics for further processing by application of rules based logic to produce production quality video for videoconferencing is described. Sensory devices and equipment, for example motion, infrared, audio, sound source localization (SSL) and video are used to detect room activity or room stimulus. The room activity is analyzed (for example, to determine whether individuals are in the subject room, speaker identification and movement within the room) and processed to produce room analytics. Speaker identification and relative placement information are used to logically depict speakers conversing with one another.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.