Generating an image mask using machine learning

A system comprising:

• one or more processors of a machine;
• and
• a memory storing instructions that, when executed by the one or more processors, cause the machine to perform operations comprising:
• generating a plurality of images, each of the plurality of images comprising a plurality of pixels;
• generating, by the one or more processors, keyframe mask data for keyframes of the plurality of images using a machine learning scheme, the keyframe mask data assigning pixels from the plurality of pixels to a plurality of segments, each segment from the plurality of segments describing a type of image feature area depicted in the plurality of images;
• generating, by the one or more processors, non-keyframe mask data for non-keyframes of the plurality of images by, for a given non-keyframe, applying a guided filter to the given non-keyframe and a preceding keyframe mask, the preceding keyframe mask generated by the machine learning scheme from a keyframe that precedes the given non-keyframe in the plurality of images;
• generating, by the one or more processors, a plurality of modified images by applying a visual effect to pixels in the plurality of images that correspond to a specified segment from the plurality of segments;
• and
• storing, by the one or more processors, the plurality of modified images in the memory.

A non-transitory machine-readable storage device embodying instructions that, when executed by one or more processors of a machine, cause the machine to perform operations comprising:

• generating a plurality of images, each of the plurality of images comprising a plurality of pixels;
• generating keyframe mask data for keyframes of the plurality of images using a machine learning scheme, the keyframe mask data assigning pixels from the plurality of pixels to a plurality of segments, each segment from the plurality of segments describing a type of image feature area depicted in the plurality of images;
• generating non-keyframe mask data for non-keyframes of the plurality of images by, for a given non-keyframe, applying a guided filter to the given non-keyframe and a preceding keyframe mask, the preceding keyframe mask generated by the machine learning scheme from a keyframe that precedes the given non-keyframe in the plurality of images;
• generating a plurality of modified images by applying a visual effect to pixels in the one or more images that correspond to a specified segment from the plurality of segments;
• and
• storing the plurality of modified images in memory of the machine.