公开(公告)号：US20230274535A1

公开(公告)日：2023-08-31

申请号：US17680906

申请日：2022-02-25

Applicant: ADOBE INC.

Inventor： Yijun Li ,  Utkarsh Ojha ,  Richard Zhang ,  Jingwan Lu ,  Elya Shechtman ,  Alexei A. Efros

IPC: G06V10/774 ,  G06F3/04842

CPC classification number: G06V10/7747 ,  G06F3/04842

Abstract: An image generation system enables user input during the process of training a generative model to influence the model's ability to generate new images with desired visual features. A source generative model for a source domain is fine-tuned using training images in a target domain to provide an adapted generative model for the target domain. Interpretable factors are determined for the source generative model and the adapted generative model. A user interface is provided that enables a user to select one or more interpretable factors. The user-selected interpretable factor(s) are used to generate a user-adapted generative model, for instance, by using a loss function based on the user-selected interpretable factor(s). The user-adapted generative model can be used to create new images in the target domain.

公开(公告)号：US20220254071A1

公开(公告)日：2022-08-11

申请号：US17163284

申请日：2021-01-29

Applicant: Adobe Inc.

Inventor： Utkarsh Ojha ,  Yijun Li ,  Richard Zhang ,  Jingwan Lu ,  Elya Shechtman ,  Alexei A. Efros

IPC: G06T11/00 ,  G06K9/62 ,  G06N3/02

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media for accurately and efficiently modifying a generative adversarial neural network using few-shot adaptation to generate digital images corresponding to a target domain while maintaining diversity of a source domain and realism of the target domain. In particular, the disclosed systems utilize a generative adversarial neural network with parameters learned from a large source domain. The disclosed systems preserve relative similarities and differences between digital images in the source domain using a cross-domain distance consistency loss. In addition, the disclosed systems utilize an anchor-based strategy to encourage different levels or measures of realism over digital images generated from latent vectors in different regions of a latent space.

公开(公告)号：US12230014B2

公开(公告)日：2025-02-18

申请号：US17680906

申请日：2022-02-25

Applicant: ADOBE INC.

Inventor： Yijun Li ,  Utkarsh Ojha ,  Richard Zhang ,  Jingwan Lu ,  Elya Shechtman ,  Alexei A. Efros

IPC: G06V10/774 ,  G06F3/04842

Abstract: An image generation system enables user input during the process of training a generative model to influence the model's ability to generate new images with desired visual features. A source generative model for a source domain is fine-tuned using training images in a target domain to provide an adapted generative model for the target domain. Interpretable factors are determined for the source generative model and the adapted generative model. A user interface is provided that enables a user to select one or more interpretable factors. The user-selected interpretable factor(s) are used to generate a user-adapted generative model, for instance, by using a loss function based on the user-selected interpretable factor(s). The user-adapted generative model can be used to create new images in the target domain.

公开(公告)号：US11763495B2

公开(公告)日：2023-09-19

申请号：US17163284

申请日：2021-01-29

Applicant: Adobe Inc.

Inventor： Utkarsh Ojha ,  Yijun Li ,  Richard Zhang ,  Jingwan Lu ,  Elya Shechtman ,  Alexei A. Efros

IPC: G06T11/00 ,  G06N3/02 ,  G06F18/22 ,  G06F18/214

CPC classification number: G06T11/00 ,  G06F18/214 ,  G06F18/22 ,  G06N3/02

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media for accurately and efficiently modifying a generative adversarial neural network using few-shot adaptation to generate digital images corresponding to a target domain while maintaining diversity of a source domain and realism of the target domain. In particular, the disclosed systems utilize a generative adversarial neural network with parameters learned from a large source domain. The disclosed systems preserve relative similarities and differences between digital images in the source domain using a cross-domain distance consistency loss. In addition, the disclosed systems utilize an anchor-based strategy to encourage different levels or measures of realism over digital images generated from latent vectors in different regions of a latent space.