公开(公告)号：US11514632B2

公开(公告)日：2022-11-29

申请号：US17091440

申请日：2020-11-06

Applicant: Adobe Inc.

Inventor： Bryan Russell ,  Taesung Park ,  Richard Zhang ,  Junyan Zhu ,  Alexander Andonian

IPC: G09G5/00 ,  G06T11/60 ,  G06N3/04

Abstract: This disclosure describes methods, non-transitory computer readable storage media, and systems that utilize a contrastive perceptual loss to modify neural networks for generating synthetic digital content items. For example, the disclosed systems generate a synthetic digital content item based on a guide input to a generative neural network. The disclosed systems utilize an encoder neural network to generate encoded representations of the synthetic digital content item and a corresponding ground-truth digital content item. Additionally, the disclosed systems sample patches from the encoded representations of the encoded digital content items and then determine a contrastive loss based on the perceptual distances between the patches in the encoded representations. Furthermore, the disclosed systems jointly update the parameters of the generative neural network and the encoder neural network utilizing the contrastive loss.

公开(公告)号：US20220148242A1

公开(公告)日：2022-05-12

申请号：US17091440

申请日：2020-11-06

Applicant: Adobe Inc.

Inventor： Bryan Russell ,  Taesung Park ,  Richard Zhang ,  Junyan Zhu ,  Alexander Andonian

IPC: G06T11/60 ,  G06N3/04

Abstract: This disclosure describes methods, non-transitory computer readable storage media, and systems that utilize a contrastive perceptual loss to modify neural networks for generating synthetic digital content items. For example, the disclosed systems generate a synthetic digital content item based on a guide input to a generative neural network. The disclosed systems utilize an encoder neural network to generate encoded representations of the synthetic digital content item and a corresponding ground-truth digital content item. Additionally, the disclosed systems sample patches from the encoded representations of the encoded digital content items and then determine a contrastive loss based on the perceptual distances between the patches in the encoded representations. Furthermore, the disclosed systems jointly update the parameters of the generative neural network and the encoder neural network utilizing the contrastive loss.