公开(公告)号：US12299896B1

公开(公告)日：2025-05-13

申请号：US17702202

申请日：2022-03-23

Applicant: Amazon Technologies, Inc.

Inventor： Qianli Feng ,  Raghu Deep Gadde ,  Pietro Perona ,  Aleix Margarit Martinez

IPC: G06T3/14 ,  G06N3/045 ,  G06T7/187 ,  G06V10/74 ,  G06V10/774

Abstract: Described herein is a computer-implemented method for generating a synthetic image. An input image can be received by a computing device. A representation of the input image on an image approximation manifold can be identified by inputting the input image into a machine learning model. The image approximation manifold can be defined by the machine learning model. A local region of the image approximation manifold can be modified relative to the first representation to generate a modified image approximation manifold. The modified image approximation manifold can include a second representation of the input image. A synthetic image can be generated based on the modified image approximation manifold. A rendering of the synthetic image can be caused on a display.

公开(公告)号：US11238596B1

公开(公告)日：2022-02-01

申请号：US16369908

申请日：2019-03-29

Applicant: Amazon Technologies, Inc.

Inventor： Raghu Deep Gadde ,  Peter Vincent Gehler

IPC: G06T7/194 ,  G06K9/72 ,  G06K9/62

Abstract: Devices and techniques are generally described for object co-segmentation in image data using a semantic filter. In various examples, first image data and second image data may be received. First semantic feature data may be determined for the first image data and second semantic feature data may be determined for the second image data. Filtered semantic feature data corresponding to the first semantic feature data may be generated by filtering the first semantic feature data and the second semantic feature data in a semantic feature space. A determination may be made that a first pixel corresponds to a first class of an object, where at least one object of the first class is represented in the first image data and the second image data.

公开(公告)号：US11568576B1

公开(公告)日：2023-01-31

申请号：US17117784

申请日：2020-12-10

Applicant: Amazon Technologies, Inc.

Inventor： Aleix Margarit Martinez ,  Raghu Deep Gadde ,  Qianli Feng ,  Alexandru Indrei ,  Gerard Gjonej

IPC: G06T11/00 ,  G06T7/11 ,  G06K9/62 ,  G06T7/70

Abstract: Techniques are generally described for generation of photorealistic synthetic image data. A generator network generates first synthetic image data. A first class of image data represented by a first portion of the first synthetic image data is detected and the first portion is sent to a first discriminator network. The first discriminator network generates a prediction of whether the first portion of the first synthetic image data is synthetically generated. A second class of image data represented by a second portion of the first synthetic image data is detected and the second portion is sent to a second discriminator network. The second discriminator network generates a prediction of whether the second portion of the first synthetic image data is synthetically generated. The generator network is updated based on the predictions of the discriminators.

公开(公告)号：US11790558B1

公开(公告)日：2023-10-17

申请号：US17363504

申请日：2021-06-30

Applicant: Amazon Technologies, Inc.

Inventor： Guha Balakrishnan ,  Raghu Deep Gadde ,  Pietro Perona ,  Aleix Margarit Martinez

IPC: G06N20/00 ,  G06T7/00 ,  G06V10/75 ,  G06F18/214

CPC classification number: G06T7/97 ,  G06F18/214 ,  G06N20/00 ,  G06V10/76

Abstract: Techniques are generally described for generation of synthetic image data. In some examples, a selection of a first image may be received. The first image may depict at least a first object having a plurality of image attributes representing visual characteristics of the at least the first object. In some examples, a selection of a first image attribute of the plurality of image attributes to be maintained in subsequently-generated images may be received. In various examples, a first machine learning model may generate a second image having the plurality of image attributes. The change in an appearance of the first image attribute may be minimized in the second image while a change in the appearance of other attributes of the plurality of image attributes may be maximized in the second image.