Amazon Technologies, Inc.

Assaf Neuberger

Arik Poznanski

Eduard Oks

Sharon Alpert

Alexander Lorbert

Bar Hilleli

Nicole C Bishop

Systems and methods are described for programmatic generation of object images with polygonal outlines. Programmatic generation can include obtaining a segmentation mask identifying pixels of a source image depicting an object, applying morphological dilation the segmentation mask to produce a dilated mask, applying a polygonal approximation algorithm to one or more contours within the dilation mask to produce a polygonal outline mask, and generating a synthetic image of the object on a designated background by overlaying a portion of the source image, designated by the segmentation mask, on to a portion of the designated background designated by the polygonal outline mask.

Programmatic generation of object images with polygonal outlines

First image data representing a first human wearing a first article of clothing may be received. The first image data, when rendered on a display, may include a first photometric artifact. A first generator network may be used to generate second image data from the first image data. The first photometric artifact may be removed from the second image data. A second generator network may be used to generate third image data from the second image data, the third image data representing the first human in a different pose relative to the first image data. Fourth image data representing the first article of clothing segmented from the first human may be generated and displayed on a display.

Catalog normalization and segmentation for fashion images

Growing community of inventors

Ness Ziona, Israel

Assaf Neuberger