US 12,354,601 B1
Using artificial creators to generate media content
Ayman Mahfouz, Culver City, CA (US); Pragyana K. Mishra, Seattle, WA (US); and Sanjeev Kumar, Redmond, WA (US)
Assigned to Amazon Technologies, Inc., Seattle, WA (US)
Filed by Amazon Technologies, Inc., Seattle, WA (US)
Filed on Dec. 10, 2021, as Appl. No. 17/547,927.
Int. Cl. G10L 15/00 (2013.01); G06N 20/20 (2019.01); G10L 13/10 (2013.01); G10L 15/18 (2013.01); H04N 21/439 (2011.01)
CPC G10L 15/1807 (2013.01) [G06N 20/20 (2019.01); G10L 13/10 (2013.01); H04N 21/4394 (2013.01)] 17 Claims
OG exemplary drawing
 
1. A first computer system comprising at least one data store and at least one computer processor,
wherein the first computer system is connected to one or more networks, and
wherein the at least one data store has one or more sets of instructions stored thereon that, when executed by the at least one computer processor, cause the first computer system to perform a method comprising:
receiving first media content from a second computer system associated with a creator of a media program, wherein the first media content represents at least a first set of words;
receiving second media content from a third computer system associated with a media source;
causing at least the first media content and the second media content to be transmitted to at least a fourth computer system associated with a listener in accordance with a media program;
processing at least the first media content to identify at least the first set of words;
determining a topic of the media program based at least in part on the first set of words;
determining information regarding at least one of a prosody of the creator or an acoustic characteristic of the creator based at least in part on the first media content;
identifying at least a first media entity expressed in the second media content;
training at least a first machine learning model using at least the first set of words, the topic, the first media entity, and the information regarding the at least one of the prosody of the creator or the acoustic characteristic of the creator;
selecting, by the first machine learning model, at least one of a second set of words or a second media entity based at least in part on at least one output received from the first machine learning model; and
causing at least third media content to be transmitted to at least the fourth computer system in accordance with the media program,
wherein the third media content represents the at least one of the second set of words or the second media entity.