Systems and methods for encoding videos using reference objects

A system comprising:

• an identification module, stored in memory, that identifies, by a computing device, a video to be encoded;
• an object module, stored in memory, that identifies, by the computing device, a set of objects that appear in the video as reference images for video encoding;
• a training module, stored in memory, that trains a machine learning algorithm to detect an object from the set of objects;
• an encoding module, stored in memory, that encodes each frame of the video using the trained machine learning algorithm, wherein the encoding module encodes each frame of the video by:
• detecting at least one object in each frame using the machine learning algorithm;
• compressing each frame to reduce data on non-object images;
• and
• performing reference-frame compression for frames in which no object is detected by the machine learning algorithm,
• wherein the reference-frame compression comprises:
• selecting at least one reference frame;
• comparing at least one later frame of the video with the reference frame;
• and
• compressing the later frame to reduce redundant image data;
• and
• at least one processor that executes the identification module, the object module, the training module, and the encoding module.

A non-transitory computer-readable medium comprising one or more computer-executable instructions that, when executed by at least one processor of a computing device, cause the computing device to:

• identify a video to be encoded;
• identify a set of objects that appear in the video as reference images for video encoding;
• train a machine learning algorithm to detect an object from the set of objects;
• and
• encode each frame of the video using the trained machine learning algorithm, wherein the instructions cause the at least one processor to encode each frame of the video at least in part by:
• detecting at least one object in each frame using the machine learning algorithm;
• compressing each frame to reduce data on non-object images;
• and
• performing reference-frame compression for frames in which no object is detected by the machine learning algorithm, wherein the reference-frame compression comprises:
• selecting at least one reference frame;
• comparing at least one later frame of the video with the reference frame;
• and
• compressing the later frame to reduce redundant image data.