Abstract: Systems and methods for information retrieval are described. Embodiments generate a dense embedding for each of a plurality of media objects to be searched, generate a sparse embedding for each of the media objects using an encoder that takes the dense embedding as an input, wherein the sparse embedding satisfies a sparsity constraint that is applied to at least one layer of the encoder during training, and perform a search on the plurality of media objects based at least in part on the sparse embedding.

Abstract: Disclosed are systems and methods for dynamically determining categories for images. A computer-implemented method may include training a neural network to receive an input image and determine one or more image categories associated with the input image; obtaining a set of images associated with a user; determining, using the trained neural network, one or more image categories associated with each image included in the obtained set of images; determining one or more dominant image categories associated with the user based on the determined image categories for the obtained set of images; and determining an image editing user interface for the user based on the determined one or more dominant image categories.

Abstract: The present disclosure relates to an image merging system that automatically and seamlessly detects and merges missing people for a set of digital images into a composite group photo. For instance, the image merging system utilizes a number of models and operations to automatically analyze multiple digital images to identify a missing person from a base image, segment the missing person from the second image, and generate a composite group photo by merging the segmented image of the missing person into the base image. In this manner, the image merging system automatically creates merged group photos that appear natural and realistic.

Abstract: Disclosed are systems and methods for dynamically determining categories for images. A computer-implemented method may include training a neural network to receive an input image and determine one or more image categories associated with the input image; obtaining a set of images associated with a user; determining, using the trained neural network, one or more image categories associated with each image included in the obtained set of images; determining one or more dominant image categories associated with the user based on the determined image categories for the obtained set of images; and determining an image editing user interface for the user based on the determined one or more dominant image categories.

Abstract: Disclosed are systems and methods for dynamically determining categories for images. A computer-implemented method includes training a neural network to receive an input image and determine one or more image categories associated with the input image; obtaining a set of images associated with a user; and determining, using the trained neural network, one or more image categories associated with each image included in the obtained set of images.

Abstract: Systems and methods for information retrieval are described. Embodiments generate a dense embedding for each of a plurality of media objects to be searched, generate a sparse embedding for each of the media objects using an encoder that takes the dense embedding as an input, wherein the sparse embedding satisfies a sparsity constraint that is applied to at least one layer of the encoder during training, and perform a search on the plurality of media objects based at least in part on the sparse embedding.

Abstract: Disclosed are systems and methods for dynamically determining categories for images. A computer-implemented method includes training a neural network to receive an input image and determine one or more image categories associated with the input image; obtaining a set of images associated with a user; and determining, using the trained neural network, one or more image categories associated with each image included in the obtained set of images.

Abstract: The present disclosure relates to an image merging system that automatically and seamlessly detects and merges missing people for a set of digital images into a composite group photo. For instance, the image merging system utilizes a number of models and operations to automatically analyze multiple digital images to identify a missing person from a base image, segment the missing person from the second image, and generate a composite group photo by merging the segmented image of the missing person into the base image. In this manner, the image merging system automatically creates merged group photos that appear natural and realistic.

Abstract: Techniques for optimizing the performance of a webpage crawler are described. According to various embodiments, historical web crawler performance data is accessed, the data describing a performance of a web crawler during various time periods in one or more prior days. A capacity of the web crawler to fulfill uniform resource locator (URL) crawl requests for an upcoming given time period is then estimated, based on the historical web crawler performance data. Thereafter, a plurality of URL crawl requests are distributed to the web crawler during the upcoming given time period, based on the estimated capacity of the web crawler.

Abstract: Techniques for optimizing the performance of a webpage crawler are described. According to various embodiments, historical web crawler performance data is accessed, the data describing a performance of a web crawler during various time periods in one or more prior days. A capacity of the web crawler to fulfil uniform resource locator (URL) crawl requests for an upcoming given time period is then estimated, based on the historical web crawler performance data. Thereafter, a plurality of URL crawl requests are distributed to the web crawler during the upcoming given time period, based on the estimated capacity of the web crawler.

Abstract: Techniques for providing improved product suggestions for related items are described. According to various embodiments, a product listing webpage associated with a retailer website that describes a specific product may be crawled, the specific product being included in a product inventory of the retailer website. Thereafter, product relations information associated with the specific product may be identified in the product listing webpage, the product relations information describing a group of one or more additional products in the product inventory having a particular relationship with the specific product. The product relations information may then be transposed to a second product inventory associated with a second retailer.

Abstract: Techniques for competitive pricing analysis and inventory management are described. According to various exemplary embodiments, a competitive pricing system is configured to crawl competitor websites for comparative pricing information at various time intervals. Moreover, the competitive pricing system is configured to determine if a price for an item on a home retailer website represents a “deal”, based on information crawled from competitor websites. According to various exemplary embodiments, a managed inventory repository system may enable improved identification of deals and specials within an inventory of a retailer website. For example, the managed inventory repository system may perform data mining operations to identify deals or specials offered for inventory items on a home retailer website. In some embodiments, a “special” can be defined in a variety of ways to suit different business units, campaigns, metrics, etc.

Abstract: Embodiments for obtaining size and brand information for a plurality of descriptors that include item types and that are associated with user profiles. The descriptors, size, and brand information are obtained by crowdsourcing and by data mining transaction data. Low confidence machine learned data may be boosted by crowdsourcing through targeted questions. Co-occurrences among descriptors are determined and categorized. Signal strength and confidence scores are calculated for the co-occurrences. Relationships between sizes and brands for the item types are calculated and confidence factors for the relationships are calculated.