Abstract: Techniques for media processing are described. A system may initiate and manage a communication session between two or more devices using a media session orchestrator (MESO). The MESO interacts with components associated with media processing services, media routing services, and signaling edge services, enabling devices from one or more networks to send and receive data. During the communication session, the system may perform media processing on data being sent between devices in order to provide additional functionality. For example, video content may be processed to provide subtitles in a plurality of languages, filters may be applied to change color balance or perform colorization of black and white content, or the like. The MESO architecture comprises a plurality of MESO routers and MESO executors, with the MESO routers being configured to route triggers and events associated with a specific media session to a single MESO executor that processes the media session.

Abstract: Methods and devices for causing a communications session between a first device and a second device to end based on lack of speech activity are described herein. In some embodiments, a communications between a first device and a second device may be initiated by the first device, where a first user account associated with the first device is authorized to initiate communications session with the second device by a second user account. After the communications session is started, audio data is received by a speech activity detection system, which determines whether the audio data represents speech or non-speech. If, after the communications session begins, non-speech is detected by the first device for more than a predefined amount of time, then the communications session is caused to end so that the first device is not capable of receiving video and/or audio associated with the second device.