Adobe, Inc.

Google Inc.

John Thomas Nack

Elya Shechtman

Jingwan Lu

Shabnam Ghadar

Ratheesh Kalarot

Baldo Antonio Faieta

Wei-An Lin

Timothy M Converse

Cameron Smith

Sevket Derin Babacan

Akhilesh Kumar

Krzysztof Kulewski

Hui Qu

Juan Carlos Miguel Anorga

Christoph Moskalonek

Jordi Miguel Verdu Orts

Kevin Duarte

An item recommendation system receives a set of recommendable items and a request to select, from the set of recommendable items, a contrast group. The item recommendation system selects a contrast group from the set of recommendable items by applying a image modification model to the set of recommendable items. The image modification model includes an item selection model configured to determine an unbiased conversion rate for each item of the set of recommendable items and select a recommended item from the set of recommendable items having a greatest unbiased conversion rate. The image modification model includes a contrast group selection model configured to select, for the recommended item, a contrast group comprising the recommended item and one or more contrast items. The item recommendation system transmits the contrast group responsive to the request.

Semantic-aware initial latent code selection for text-guided image editing and generation

Systems and methods for image processing are described. Embodiments of the present disclosure encode features of a source image to obtain a source appearance encoding that represents inherent attributes of a face in the source image; encode features of a target image to obtain a target non-appearance encoding that represents contextual attributes of the target image; combine the source appearance encoding and the target non-appearance encoding to obtain combined image features; and generate a modified target image based on the combined image features, wherein the modified target image includes the inherent attributes of the face in the source image together with the contextual attributes of the target image.

Disentangling latent representations for image reenactment

The present disclosure describes systems and methods for image processing. Embodiments of the present disclosure include an image processing apparatus configured to generate modified images (e.g., synthetic faces) by conditionally changing attributes or landmarks of an input image. A machine learning model of the image processing apparatus encodes the input image to obtain a joint conditional vector that represents attributes and landmarks of the input image in a vector space. The joint conditional vector is then modified, according to the techniques described herein, to form a latent vector used to generate a modified image. In some cases, the machine learning model is trained using a generative adversarial network (GAN) with a normalization technique, followed by joint training of a landmark embedding and attribute embedding (e.g., to reduce inference time).

High resolution conditional face generation

Implementations relate to computer-assisted cropping of an image. A computer-executed method includes receiving user input indicative of a change in size of a crop window from a first size to a second size, where the crop window is displayed over an image in an image editing user interface. The method detects that the crop window of the second size has an aspect ratio within a threshold range of a predetermined aspect ratio. In response to the detection, the method resizes the crop window to a third size, where the crop window of the third size has a resized aspect ratio substantially the same as the predetermined aspect ratio.

Computer-assisted image cropping

A method including receiving input identifying a panel associated with a software application, the software application including a Graphical User Interface (GUI) and the panel comprising a discrete area of the GUI within which to group software objects to control the software application. Additionally, the method includes processing the input to identify an object of the software objects to be included in the panel. Further, the method includes generating a panel definition including the object and dimensions of the panel, the panel definition to be processed to render the GUI.

Panel configurator engine

A virtual canvas is displayed at a graphical user interface of a touch-enabled device, a virtual canvas. A plurality of color indicia is displayed. A first input that selects a first selected color from the plurality of color indicia is received. A first positional input indicating a first position at which the first selected color is displayed is also received. A second input that selects a second selected color from the plurality of color indicia is received. A second positional input indicating a second position at which the second selected color is displayed is received at the virtual canvas. A mixed color based on first selected color and the second selected color is generated if the first position overlaps the second position. The mixed color is sent from the touch-enabled device to a computing device through a network.

Remote paint mixing controller

Growing community of inventors

San Jose, CA, United States of America

John Thomas Nack