Methods and systems for interfacing heterogeneous endpoints and web-based media sources in a video conference

摘要

Described herein are methods and systems for interfacing heterogeneous endpoints and web-based media sources in a video conference. Interfacing heterogeneous endpoints can be challenging due to audio/video (A/V) streams being encoded in a format that is proprietary to a media provider of the A/V stream (e.g., format that is proprietary to Skype). To address such a challenge, a client application from the media provider may be installed at a virtual machine, and may be used to decode A/V streams that are encoded in a format proprietary to the media provider. Further, to display content from a web-based media source (e.g., YouTube) in a video conference, a browser may be installed at a virtual machine, and may be used to virtually render a media signal from the web-based media source, such virtually rendered media signal subsequently being captured and presented in a video conference.

主权项

1. A method for processing audio/video (A/V) streams of a video conference at one or more servers, the method comprising:
for each of a plurality of user endpoints of the video conference:
receiving an A/V stream, the A/V stream encoded in a format that is proprietary to a media provider of the A/V stream; andtranscoding the A/V stream by (i) decoding the A/V stream using a client application from the media provider, the client application executing at a virtual machine instantiated on the one or more servers, and (ii) encoding the decoded A/V stream into a format that is compatible with a media mixer, wherein each of the virtual machines instantiated on the one or more servers only executes a single client application, and wherein load prediction is employed to minimize a number of virtual machines that are instantiated on the one or more servers at any time instance, the load prediction comprising:
predicting a first number of client applications that will be needed at a first time instance to decode the A/V streams received at the one or more servers;before the first time instance, (i) deploying the first number of virtual machines on the one or more servers, and (ii) installing the first number of client applications, one on each of the deployed virtual machines;at the first time instance, determining whether an additional number of client applications are needed to decode the A/V streams, the additional number of client applications in excess of the first number of client applications; andif an additional number of client applications are needed, (i) deploying a corresponding number of additional virtual machines on the one or more servers, and (ii) installing the additional number of client applications, one on each of the additional virtual machines.