Abstract: In some embodiments, a context-based translation application generates a co-occurrence data structure for a target language describing co-occurrences of target language words and source language words. The context-based translation application receives an input tag for an input image in the source language to be translated into the target language. The context-based translation application obtains multiple candidate translations in the target language for the input tag and determines a translated tag from the multiple candidate translations based on the co-occurrence data structure. The context-based translation application further associates the translated tag with the input image.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media that implement an inpainting framework having computer-implemented machine learning models to generate high-resolution inpainting results. For instance, in one or more embodiments, the disclosed systems generate an inpainted digital image utilizing a deep inpainting neural network from a digital image having a replacement region. The disclosed systems further generate, utilizing a visual guide algorithm, at least one deep visual guide from the inpainted digital image. Using a patch match model and the at least one deep visual guide, the disclosed systems generate a plurality of modified digital images from the digital image by replacing the region of pixels of the digital image with replacement pixels. Additionally, the disclosed systems select, utilizing an inpainting curation model, a modified digital image from the plurality of modified digital images to provide to a client device.

Abstract: Systems and methods for product retrieval are described. One or more aspects of the systems and methods include receiving a query that includes a text description of a product associated with a brand; identifying the product based on the query by comparing the text description to a product embedding of the product, wherein the product embedding is based on a brand embedding of the brand; and displaying product information for the product in response to the query, wherein the product information includes the brand.

Abstract: Systems and methods for image processing are described. One or more embodiments of the present disclosure identify a latent vector representing an image of a face, identify a target attribute vector representing a target attribute for the image, generate a modified latent vector using a mapping network that converts the latent vector and the target attribute vector into a hidden representation having fewer dimensions than the latent vector, wherein the modified latent vector is generated based on the hidden representation, and generate a modified image based on the modified latent vector, wherein the modified image represents the face with the target attribute.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media that generate inpainted digital images utilizing class-specific cascaded modulation inpainting neural network. For example, the disclosed systems utilize a class-specific cascaded modulation inpainting neural network that includes cascaded modulation decoder layers to generate replacement pixels portraying a particular target object class. To illustrate, in response to user selection of a replacement region and target object class, the disclosed systems utilize a class-specific cascaded modulation inpainting neural network corresponding to the target object class to generate an inpainted digital image that portrays an instance of the target object class within the replacement region.

Abstract: Systems and methods for image processing are described. Embodiments identify a training set including a first image that includes a ground truth blur classification and second image that includes a ground truth blur map, generate a first embedded representation of the first image and a second embedded representation of the second image using an image encoder, predict a blur classification of the first image based on the first embedded representation using a classification layer, predict a blur map of the second image based on the second embedded representation using a map decoder, compute a classification loss based on the predicted blur classification and the ground truth blur classification, train the image encoder and the classification layer based on the classification loss, compute a map loss based on the blur map and the ground truth blur map, and train the image encoder and the map decoder.

Abstract: Embodiments of the present invention provide an automated image tagging system that can predict a set of tags, along with relevance scores, that can be used for keyword-based image retrieval, image tag proposal, and image tag auto-completion based on user input. Initially, during training, a clustering technique is utilized to reduce cluster imbalance in the data that is input into a convolutional neural network (CNN) for training feature data. In embodiments, the clustering technique can also be utilized to compute data point similarity that can be utilized for tag propagation (to tag untagged images). During testing, a diversity based voting framework is utilized to overcome user tagging biases. In some embodiments, bigram re-weighting can down-weight a keyword that is likely to be part of a bigram based on a predicted tag set.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media that generate inpainted digital images utilizing a cascaded modulation inpainting neural network. For example, the disclosed systems utilize a cascaded modulation inpainting neural network that includes cascaded modulation decoder layers. For example, in one or more decoder layers, the disclosed systems start with global code modulation that captures the global-range image structures followed by an additional modulation that refines the global predictions. Accordingly, in one or more implementations, the image inpainting system provides a mechanism to correct distorted local details. Furthermore, in one or more implementations, the image inpainting system leverages fast Fourier convolutions block within different resolution layers of the encoder architecture to expand the receptive field of the encoder and to allow the network encoder to better capture global structure.

Abstract: Certain embodiments involve a method for generating a search result. The method includes processing devices performing operations including receiving a query having a text input by a joint embedding model trained to generate an image result. Training the joint embedding model includes accessing a set of images and textual information. Training further includes encoding the images into image feature vectors based on spatial features. Further, training includes encoding the textual information into textual feature vectors based on semantic information. Training further includes generating a set of image-text pairs based on matches between image feature vectors and textual feature vectors. Further, training includes generating a visual grounding dataset based on spatial information. Training further includes generating a set of visual-semantic joint embeddings by grounding the image-text pairs with the visual grounding dataset.

Abstract: Systems and methods for image processing are configured. Embodiments of the present disclosure encode a content image and a style image using a machine learning model to obtain content features and style features, wherein the content image includes a first object having a first appearance attribute and the style image includes a second object having a second appearance attribute; align the content features and the style features to obtain a sparse correspondence map that indicates a correspondence between a sparse set of pixels of the content image and corresponding pixels of the style image; and generate a hybrid image based on the sparse correspondence map, wherein the hybrid image depicts the first object having the second appearance attribute.

Abstract: The systems, methods, a non-transitory computer readable mediums relate to an object selection system that accurately detects and automatically selects user-requested objects (e.g., query objects) in a digital image. For example, the object selection system builds and utilizes an object selection pipeline to determine which object detection neural network to utilize to detect a query object based on analyzing the object class of the query object. In addition, the object selection system can add, update, or replace portions of the object selection pipeline to improve overall accuracy and efficiency of automatic object selection within an image.

Abstract: Systems and methods for image tagging are described. In some embodiments, images with problematic tags are identified after applying an auto-tagger. The images with problematic tags are then sent to an object detection network. In some cases, the object detection network is trained using a training set selected to improve detection of objects associated with the problematic tags. The output of the object detection network can be merged with the output of the auto-tagger to provide a combined image tagging output. In some cases, the output of the object detection network also includes a bounding box, which can be used to crop the image around a relevant object so that the auto-tagger can be reapplied to a portion of the image.

Abstract: In implementations of resource-aware training for neural network, one or more computing devices of a system implement an architecture optimization module for monitoring parameter utilization while training a neural network. Dead neurons of the neural network are identified as having activation scales less than a threshold. Neurons with activation scales greater than or equal to the threshold are identified as survived neurons. The dead neurons are converted to reborn neurons by adding the dead neurons to layers of the neural network having the survived neurons. The reborn neurons are prevented from connecting to the survived neurons for training the reborn neurons.

Abstract: A group captioning system includes computing hardware, software, and/or firmware components in support of the enhanced group captioning contemplated herein. In operation, the system generates a target embedding for a group of target images, as well as a reference embedding for a group of reference images. The system identifies information in-common between the group of target images and the group of reference images and removes the joint information from the target embedding and the reference embedding. The result is a contrastive group embedding that includes a contrastive target embedding and a contrastive reference embedding with which to construct a contrastive group embedding, which is then input to a model to obtain a group caption for the target group of images.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media that generates composite images via auto-compositing features. For example, in one or more embodiments, the disclosed systems determine a background image and a foreground object image for use in generating a composite image. The disclosed systems further provide, for display within a graphical user interface of a client device, at least one selectable option for executing an auto-composite model for the composite image, the auto-composite model comprising at least one of a scale prediction model, a harmonization model, or a shadow generation model. The disclosed systems detect, via the graphical user interface, a user selection of the at least one selectable option and generate, in response to detecting the user selection, the composite image by executing the auto-composite model using the background image and the foreground object image.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media that utilizes artificial intelligence to learn to recommend foreground object images for use in generating composite images based on geometry and/or lighting features. For instance, in one or more embodiments, the disclosed systems transform a foreground object image corresponding to a background image using at least one of a geometry transformation or a lighting transformation. The disclosed systems further generating predicted embeddings for the background image, the foreground object image, and the transformed foreground object image within a geometry-lighting-sensitive embedding space utilizing a geometry-lighting-aware neural network. Using a loss determined from the predicted embeddings, the disclosed systems update parameters of the geometry-lighting-aware neural network. The disclosed systems further provide a variety of efficient user interfaces for generating composite digital images.

Abstract: The present disclosure relates to systems, non-transitory computer-readable media, and methods for utilizing machine learning models to generate refined depth maps of digital images utilizing digital segmentation masks. In particular, in one or more embodiments, the disclosed systems generate a depth map for a digital image utilizing a depth estimation machine learning model, determine a digital segmentation mask for the digital image, and generate a refined depth map from the depth map and the digital segmentation mask utilizing a depth refinement machine learning model. In some embodiments, the disclosed systems generate first and second intermediate depth maps using the digital segmentation mask and an inverse digital segmentation mask and merger the first and second intermediate depth maps to generate the refined depth map.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media that utilizes artificial intelligence to learn to recommend foreground object images for use in generating composite images based on geometry and/or lighting features. For instance, in one or more embodiments, the disclosed systems transform a foreground object image corresponding to a background image using at least one of a geometry transformation or a lighting transformation. The disclosed systems further generating predicted embeddings for the background image, the foreground object image, and the transformed foreground object image within a geometry-lighting-sensitive embedding space utilizing a geometry-lighting-aware neural network. Using a loss determined from the predicted embeddings, the disclosed systems update parameters of the geometry-lighting-aware neural network. The disclosed systems further provide a variety of efficient user interfaces for generating composite digital images.

Abstract: Techniques for identity preserved controllable facial image manipulation are described that support generation of a manipulated digital image based on a facial image and a render image. For instance, a facial image depicting a facial representation of an individual is received as input. A feature space including an identity parameter and at least one other visual parameter is extracted from the facial image. An editing module edits one or more of the visual parameters and preserves the identity parameter. A renderer generates a render image depicting a morphable model reconstruction of the facial image based on the edit. The render image and facial image are encoded, and a generator of a neural network is implemented to generate a manipulated digital image based on the encoded facial image and the encoded render image.

Abstract: Systems, methods, and software are described herein for removing people distractors from images. A distractor mitigation solution implemented in one or more computing devices detects people in an image and identifies salient regions in the image. The solution then determines a saliency cue for each person and classifies each person as wanted or as an unwanted distractor based at least on the saliency cue. An unwanted person is then removed from the image or otherwise reduced from the perspective of being an unwanted distraction.