Adobe, Inc.

International Business Machines Corporation

Innography, Inc.

Ryan Rozich

Sachin Soni

Jonathan Roeder

Ajay Jain

Anthony Richard Hagale

Sanjeev Tagra

Prasenjit Mondal

Paul Stuart Williamson

Jason Ernest Kelley

Nikaash Puri

Arshiya Aggarwal

Ajay Bedi

Nitish Maurya

Baldo Antonio Faieta

Ajinkya Gorakhnath Kale

Tyron Jerrod Stading

Akram Bou-Ghannam

Saeid Motiian

Michele Saad

Srinivas Hasti

John Williams Miller

Venkata Naveen Kumar Yadav Marri

Roland Geisler

Roji John

Mikhail Kotov

Dylan Nathaniel Warnock

An image search system uses a multi-modal model to determine relevance of images to a spoken query. The multi-modal model includes a spoken language model that extracts features from spoken query and a language processing model that extract features from an image. The multi-model model determines a relevance score for the image and the spoken query based on the extracted features. The multi-modal model is trained using a curriculum approach that includes training the spoken language model using audio data. Subsequently, a training dataset comprising a plurality of spoken queries and one or more images associated with each spoken query is used to jointly train the spoken language model and an image processing model to provide a trained multi-modal model.

Spoken query processing for image search

Techniques are disclosed for generating image recommendations to facilitate the sale of a product. An example methodology includes identifying a product category associated with an image of the product provided by the seller, and a product sub-category associated with the product image. The method further includes retrieving one or more images of for-sale items. The retrieval is based on a search of for-sale listings using the identified product category and the identified product sub-category. The method further includes clustering the retrieved images of for-sale items into groups, each group associated with a perspective viewpoint of the for-sale item. The method further includes providing a selected image from each group as an image recommendation. The selection is based on a value score associated with each of the images of the for-sale items. A graphical status indicating completeness of the seller's image set is updated in response to recommended images being adopted.

Generation of recommendations for visual product details

An image differentiation system receives input feature vectors for multiple input images and reference feature vectors for multiple reference images. In some cases, the feature vectors are extracted by an image feature extraction module trained based on training image triplets. A differentiability scoring module determines a differentiability score for each input image based on a distance between the input feature vectors and the reference feature vectors. The distance for each reference feature vector is modified by a weighting factor based on interaction metrics associated with the corresponding reference image. In some cases, an input image is identified as a differentiated image based on the corresponding differentiability score. Additionally or alternatively, an image modification module determines an image modification that increases the differentiability score of the input image. The image modification module generates a recommended image by applying the image modification to the input image.

Machine learning techniques for differentiability scoring of digital images

Digital image selection techniques are described that employ machine learning to select a digital image of an object from a plurality of digital images of the object. The plurality of digital images each capture the object for inclusion as part of generating digital content, e.g., a webpage, a thumbnail to represent a digital video, and so on. In one example, digital image selection techniques are described that employ machine learning to select a digital image of an object from a plurality of digital images of the object. As a result, the service provider system may select a digital image of an object from a plurality of digital images of the object that has an increased likelihood of achieving a desired outcome and may address the multitude of different ways in which an object may be presented to a user.

Machine learning for digital image selection across object variations

Embodiments of the technology described herein, provide improved visual search results by combining a visual similarity and a textual similarity between images. In an embodiment, the visual similarity is quantified as a visual similarity score and the textual similarity is quantified as a textual similarity score. The textual similarity is determined based on text, such as a title, associated with the image. The overall similarity of two images is quantified as a weighted combination of the textual similarity score and the visual similarity score. In an embodiment, the weighting between the textual similarity score and the visual similarity score is user configurable through a control on the search interface. In one embodiment, the aggregate similarity score is the sum of a weighted visual similarity score and a weighted textual similarity score.

Text adjusted visual search

Certain embodiments involve visually augmenting images of three-dimensional containers with virtual elements that fill one or more empty regions of the three-dimensional containers. For instance, a computing system receives a first image that depicts a storage container and identify sub-containers within the storage container. The computing system selects, from a virtual object library, a plurality of virtual objects that are semantically related to the sub-container. The computing system determines an arrangement of the virtual objects within the sub-container based on semantics associated with the sub-container and the plurality of virtual objects. The computing system generates a second image that depicts the arrangement of the plurality of virtual objects within the storage container and sub-containers. The computing system generates, for display, the second image depicting the storage container and the arrangement of the virtual objects.

Visually augmenting images of three-dimensional containers with virtual elements

Methods, apparatus, and systems are provided for obfuscating facial identity in images by synthesizing a new facial image for an input image. A base face is detected from or selected for an input image. Facial images that are similar to the base face are selected and combined to create a new facial image. The new facial image is added to the input image such that the input image includes a combination of the base face and the new facial image. Where no base face is detected in the input image, a base face is selected from reference facial images based at least on pose keypoints identified in the input image. After a new facial image is generated based on the selected base face, a combination of the new facial image and the base facial image are added to the input image by aligning one or more pose keypoints.

Identity obfuscation in images utilizing synthesized faces

The present disclosure relates to systems, methods, and non-transitory computer readable media for generating target products for a product search based on gesture input received via a digital canvas. For example, the disclosed systems can utilize digital image classification models to generate product sets based on individual visual product features of digital images of products. The disclosed systems can further receive gesture input within a digital canvas indicating visual product features. In addition, the disclosed systems can compare the gesture input of the digital canvas with representative digital images of product sets generated by particular classification models to identify product sets that include the indicated visual product features. Further, the disclosed systems can provide target products from the identified product sets for display via a product search interface website.

Utilizing machine learning models to generate experience driven search results based on digital canvas gesture inputs

The present disclosure relates to a computer-implemented method for generating an enhanced image from an original image, the method including segmenting the original image into a segmented image using an artificial neural network; curve fitting the segmented image to determine boundary artifacts; removing the determined boundary artifacts to generate a smoothed boundary image; and generating the enhanced image from the original image and the smoothed boundary image. The image maybe enhanced further by correcting for glare and adding artificial light.

Method and system to enhance quality of digital images

Background content suggestion for combination with identified items is described. Initially, a system receives digital visual content depicting an item, e.g., a product for purchase. The system identifies various content items that are available for suggesting as replacement backgrounds for the item and then determines a compatibility of these identified content items with the item depicted in the received digital visual content. In particular, the system determines compatibility based on both a scene compatibility and a color compatibility of the identified content items with the depicted item. Based on a combination of the scene and color compatibility, the system surfaces at least one of the identified content items (e.g., a highest scoring content item) as a suggested replacement background for the depicted item.

Background content suggestion for combination with identified items

In a particular embodiment, a method includes sending a text string from a source device to a destination device. The destination device includes a unique identifier repository having a plurality of globally unique identifiers mapped text strings, relationship data defining associations between the plurality of globally unique identifiers, and meta-data related to the text string. Each of the plurality of globally unique identifiers includes a data value that is shared by a plurality of data sources as an identifier associated with information related to the text string. The method further includes receiving data including one or more globally unique identifiers, associated relationship data, and associated meta-data from the destination device at the source device in response to sending the text string. Additionally, the method includes storing the received data at a memory of the source device for use in searching and relating search results received from multiple data sources.

Growing community of inventors

Austin, TX, United States of America

Ryan Rozich