Systems and methods for utilizing an artificial intelligence (AI) model to reduce storage resource utilization. One such method including receiving data associated with a plurality of stagnant items determined inactive for a threshold time. The method further includes, using an AI model trained using data related to the plurality of stagnant items, analyzing the data associated with the plurality of stagnant items to identify at least one candidate item of the plurality of stagnant items for including in a bundling option. The method further includes generating a bundling recommendation including the bundling option, wherein the bundling option includes the at least one candidate item. The method further includes presenting the bundling recommendation along with an acceptance indicator for accepting the bundling recommendation and a rejection indicator for rejecting the bundling recommendation.
A system for onboarding internet of things (IoT) devices. The system includes a storage device storing, for each of a plurality of IoT devices, a device profile associated with the IoT device; and an IoT edge hub comprising a processor and a computer-readable medium. The computer-readable medium stores instructions that are operative upon execution by the processor to, for each of the plurality of IoT devices: establish a low-level connection with the IoT device and receive a device identifier (ID) from the IoT device; retrieve, from the storage device, the device profile associated with the IoT device using the device ID; receive a data packet from the IoT device in a device data format associated with the IoT device; and using the device profile, convert the data packet to a converted data formatted in a predefined data format associated with the IoT edge hub.
Systems and methods for sampling item data. One such method includes receiving, from for each of a plurality of items, a recognition rate corresponding to a rate of success of correlation between: image data of the item captured by an image capture device, and an identified selection of the item from item selection data generated by an item tracker configured to generate the item selection data for lists of selected items. The method further includes determining, from the plurality of items, a plurality of target items having recognition rates failing to satisfy a recognition threshold; identifying, from the item selection data, a plurality of candidate items lists each including at least one of the plurality of target items; determining, by the computing device, a group of top items lists from the plurality of candidate items lists; and retrieving, by the computing device, image data associated with the top selected-items lists.
G06V 10/762 - Dispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique utilisant le regroupement, p. ex. de visages similaires sur les réseaux sociaux
G06F 16/583 - Recherche caractérisée par l’utilisation de métadonnées, p. ex. de métadonnées ne provenant pas du contenu ou de métadonnées générées manuellement utilisant des métadonnées provenant automatiquement du contenu
Examples may be related to anomaly detection using machine learning. An example may involve receiving a risk assessment request regarding a transaction; generating feature data based on the risk assessment request; and determining, using a machine learning model, a risk score based on the feature data. The machine learning model may be trained based on an objective function characterizing a plurality of objectives. Recommendation data regarding the transaction may be generated based on the risk score, and transmitted to a computing device.
G06Q 20/40 - Autorisation, p. ex. identification du payeur ou du bénéficiaire, vérification des références du client ou du magasinExamen et approbation des payeurs, p. ex. contrôle des lignes de crédit ou des listes négatives
Examples provide a system for converting a paper-based list of items into a digital list with greater accuracy using generative artificial intelligence (Gen AI) and historical data associated with items frequently purchased with a given geographical region. A text-based list of items is extracted from an image of a paper-based list of items using optical character recognition (OCR), optical marker recognition (OMR), and/or fuzzy matching. The list can include specific items and/or generic descriptions of item types. A machine learning (ML) model analyzes historical data for item purchasing trends and user-specific transaction data to identify the most commonly purchased specific items for each generic item type. A single predicted specific item mapped to each generic item type is used to populate a digital list representing the paper-based list with greater accuracy without user intervention. User feedback enables re-training of the ML model to further improve accuracy of the item recognition.
Examples relate to a computer-implemented method of system including a processor that can perform certain operations. The operations can include obtaining campaign inputs for a campaign associated with product specifications for an advertiser. The operations also can include generating, using an approximate nearest neighbor similarity search, targeting cuts for a targeting tactic based at least on the campaign inputs. The targeting cuts can be personalized to the advertiser and the product specifications. The operations additionally can include generating an optimized ad group plan including one or more groups of the targeting cuts and, for each of the one or more groups of the targeting cuts, a respective recommended maximum bid, a respective expected amount of winning impressions, and a respective budget allocation. The operations further can include outputting the optimized ad group plan for the campaign. Other embodiments are described.
Example systems and methods use trained models and large language models (LLMs) to generate synthetic search queries in connection with item searches. An example system includes: a database including past user search queries directed to items of interest to users; a processing resource; and a machine readable medium storing instructions that cause the processing resource to: generate, using a first trained model, metadata fields for an item where the first trained model uses past user search queries, an item classification showing a relationship of the item to other items, and/or the features and attributes of the item; generate, using a second trained model, a prompt to query LLMs to generate synthetic search queries that may be asked by a user; receive the synthetic search queries from the LLMs; and map, using a trained categorization model, each of the received synthetic search queries to item categories.
Example implementations relate to communication element selection in a network environment. In an example, a plurality of user features is received and input data derived from the plurality of user features is processed using a window logic and provided to an attention-based machine learning model. A plurality of communication elements is provided to the attention-based machine learning model, which assigns one or more weights to each of the plurality of communication elements for a subsequent time period based on the plurality of user features obtained from a preceding time period. A score for each of the plurality of communication elements based on the one or more weights for each of the plurality of communication elements is calculated and an interface including a communication element of the plurality of communication elements having a highest calculated score is generated.
A system includes an electronic database, a central hub processing resource, and a computer readable medium storing instructions that, when executed by the processing resource provide a client center user interface to receive catalog information from a client side system, electronically store the catalog information in the electronic database, and provide a fulfillment gateway. The fulfillment gateway integrates the client side system with a plurality of hub participants by making information input by the client side system accessible to the hub participants. Electronic orders are received at the hub according to the published catalog. The electronic orders are processed by the processing resource and a fulfillment option is accessed to fulfill the electronic order. The ordered item is picked and delivered to the customer.
Example implementations relate to systems and methods for detecting anomalies in performance data and changing a status based on the detected anomalies. In an example, a system receives performance data obtained during an anomaly detection window. The system determines, using a decomposer, a data decomposition of the performance data. The system determines, using an anomaly value generator; a plurality of anomaly values based on the data decomposition. The system determine, using an anomaly scorer, an anomaly score based on the plurality of anomaly values. The system also, in accordance with a determination that the anomaly score is above an anomaly threshold, generates a notification for adjusting a user status, and transmits the notification for adjusting the user status to a computing device.
Store exit verification system and method for retail purchases are provided. The system comprises a sensor array, a display device, and a control circuit. The sensor array collects information from items in a shopping container placed in a shopping container placement area. The control circuit identifies a transaction identifier associated with the shopping container, retrieves an item list associated with the transaction identifier, identifies items in the shopping container based on the information collected via the sensor array, determines whether an intervention condition is present based on a comparison of the item list with the identified items, and indicates a verification completion via the display device if no intervention condition is detected.
G08B 13/196 - Déclenchement influencé par la chaleur, la lumière, ou les radiations de longueur d'onde plus courteDéclenchement par introduction de sources de chaleur, de lumière, ou de radiations de longueur d'onde plus courte utilisant des systèmes détecteurs de radiations passifs utilisant des systèmes de balayage et de comparaison d'image utilisant des caméras de télévision
There are provided systems and methods that use trained models and large language models (LLMs) to generate synthetic search queries with tagged item attributes. The system includes: a database containing items with attributes and values associated with the attributes; a processing resource; and a machine readable medium storing instructions that cause the processing resource to: generate, using a first trained model, a first prompt to query an LLM to extract focused item attributes for an item; receive the focused product attributes; generate, using a second trained model, a second prompt based on the focused item attributes to query the LLM to generate synthetic search queries; tag, using the second trained model, each of the synthetic search queries with value pairs that include a focused item attribute and a value for that attribute; and receive the tagged synthetic search queries.
Example implementations related to automated anomaly detection in time-series datasets are disclosed. In an example, a request for an automated redemption operation is received. The request includes a plurality of time-series data elements. A reconstruction error is generated for the request using a trained autoencoder model that receives the plurality of time-series data elements. The reconstruction error is representative of a difference between the plurality of time-series data elements and a reconstructed plurality of time-series data elements generated by the trained autoencoder model. In response to determining the reconstruction error is above the predetermined threshold and that the request for the automated redemption operation satisfies at least one anomaly detection rule, execution of the automated redemption operation is prevented.
Examples may be related to cross-category recommendation. An example may involve identifying an anchor item to be presented to a user via a user interface, wherein the anchor item is in a first category; and evaluating, using a machine learning model, a degree of cross-category intent of the user based at least partially on real-time interaction data of the user. The degree of cross-category intent may indicate a likelihood that the user will engage with any item in a second category that is different from the first category. An eligibility of the user to receive an item recommendation in the second category can be determined based on the degree of cross-category intent. A recommended item in the second category may be determined based on the eligibility, and presented to the user together with the anchor item in the user interface.
A quick-connect powered case holds an electrically-powered shelf module (e.g., electronic shelf label), and has a pair of power terminal arms and a connector inside that is coupled to the power terminal arms. The case may be quickly affixed to a shelf fixture that supports a product shelf and has electrically conducting brackets. The brackets provide supplied electrical power from conducting vertical shelf rails. When the power terminal arms contact the shelf brackets and the shelf module is plugged into the connector, electrical power flows from the shelf brackets, through the power terminal arms and connector, into the shelf module. For safety, the shelf fixture has an insulating cover over the shelf brackets, with openings that permit each of the power terminal arms to electrically couple with one of the shelf brackets. The power terminal arms are sufficiently thin that the product shelf lies flat on the shelf fixture.
A47B 57/42 - Meubles à tiroirs, étagères ou rayonnages caractérisés par la possibilité de régler les rayons ou les cloisons comprenant des moyens pour régler la hauteur des supports de rayons amovibles consistant en crochets s'engageant dans des ouvertures les supports de rayons étant en porte-à-faux
A47B 57/46 - Meubles à tiroirs, étagères ou rayonnages caractérisés par la possibilité de régler les rayons ou les cloisons comprenant des moyens pour régler la hauteur des supports de rayons amovibles consistant en boulons à écrous servant d'organes de raccordement les supports de rayons étant en porte-à-faux
16.
SYSTEMS AND METHODS FOR GENERATING PERSONALIZED CONTENT
In some embodiments, apparatuses and methods are provided herein useful to generate personalized content. In some embodiments, a system comprising a processing resource; a machine readable medium storing instructions that, when executed, cause the processing resource to: aggregate, session data during a client session; update, periodically and during the interaction session, one or more inference indicators associated with the client in an inference cache storage, the one or more inference indicators being determined via a trained machine learning model using the session data and historical data in a historical data database; identify a trigger event based on client interactions; retrieve, in response to the trigger event, at least one inference indicator associated with the client from the inference cache storage; and generate a personalized content for display on the client device based on the at least one inference indicator retrieved from the inference cache storage.
Example implementations related to user attrition prediction and interface generation are disclosed. In an example, time series datasets that each include interaction data points including a corresponding time stamp are received. One or more features are extracted and a time series label is generated for each time series dataset based at least in part on a gap between each of the plurality of interaction data points. An attrition prediction model is trained using the time series datasets and the corresponding time series label. The attrition prediction model generates an attrition likelihood. A user-specific time series dataset is received and a user-specific attrition likelihood is generated. An interface intervention is generated based on the user-specific attrition likelihood and instructions are transmitted that cause an interface including the interface intervention to be displayed on a user device associated with the user-specific time series dataset.
A computer-implemented method is provided that can smooth a truck flow to a destination node. Input data can be received by a truck flow planning (TFP) system, including truck demand data, truck capacity constraints, and node capacity constraints. The TFP system can include truck demand projection (TDP), truck flow optimization (TFO), and receding horizon control (RHC) components. Truck demand can be projected over a horizon. A plan for the truck flow can be generated that reduces variability while at least maintaining inventory health and can include formulating the plan as a Mixed Integer Programming (MIP) problem. The plan can be adjusted when one or more of new input data features, outputs of the RHC component, outputs of a truck planning optimization (TPO) component, or outputs of a truck load optimization (TLO) component are within thresholds. The plan, as generated and adjusted, can be based on a rolling time window.
Examples provide a smart bagging station including a halo array of sensor devices generating sensor data associated with a detection zone surrounding the smart bagging station. The smart bagging station includes barcode scanners in a curved arrangement within a recessed sensor device housing located behind a bagging device. As a user places an item into a bag, the item is automatically scanned by the barcode scanners without re-orienting the item or passing it across a scanning sensor. Other sensor devices, including RFID tag readers, cameras, and/or weight sensors are located at multiple locations above the bagging device, below the bagging device, on the sides of the bagging device, and/or behind the bagging device creating a detection zone encompassing the bagging area. Items within the detection zone are automatically identified using different types of sensor data without requiring the manual scanning for faster and more efficient checkout.
G06V 10/70 - Dispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique
A method includes receiving candidate recommended items based on an item included in routine reorders of a user. For each candidate recommended item, dimension features for each dimension are embedded as embedding vectors. The dimensions include a user preference, a department affinity, a model suitability, and an item conversion potential dimension. The embedding vectors are combined into a feature vector input to a deep neural network (DNN) model to generate a ranking score representing a likelihood that the user engages with the candidate recommended item. Dimension-specific scores are obtained from intermediate layers of the DNN model including a user preference, a department affinity, a model suitability, and an item conversion potential score, which are output by a multilayer perceptron (MLP) network applied to respective embedding vectors. A final score is derived by incorporating the ranking score with respective weighted contributions. The candidate recommended items are ranked by the final scores.
In some embodiments, apparatuses and methods are provided herein useful to provide data generation and abstraction for items of a platform. Some embodiments, system may include a database storing data associated with a plurality of items, a processing resource, and a machine readable medium storing instructions that when executed cause the processing resource to: generate, using a first trained model, a first prompt to query at least one LLM to design a series of steps to determine one or more insights associated with a target item; receive the series of steps from the at least one LLM; generate, using an additional trained model, a step specific prompt to query the at least one LLM to provide a step specific insight factor; receive the step specific insight factor; and output data to cause a display of a user device to display the target item and the insight.
Example implementations relate to output guardrails in automated systems. In an example, a user input from a user device is received by a first trained model and an initial output is generated in response to the first trained model receiving the user input from the user device. The initial output is compared to a set of target response elements, each of the target response elements include at least a portion of a machine generated utterance. A revised output is generated by applying a mitigation logic selected based on the at least one target response element and the revised output is transmitted to the user device when the output of the first trained machine learning model includes at least one target response element. The initial output is transmitted to the user device when the initial output does not include at least one target element.
There are provided systems and methods for communicating with transaction processors and handling error codes received from them. A system may include: database(s) including records with acquisition completion methods; a processing resource; and a machine readable medium storing instructions. The processing resource may: access a record; retrieve an acquisition completion method; transmit a first communication including a first electronic acquisition completion attempt with the first acquisition completion method to an acquisition completion processor; receive an electronic communication from the acquisition completion processor including a decline indication and including an error code; determine that the error code matches one of a first set of error codes that are terminal error codes or matches one of a second set of error codes that are retry error codes. The processing resource updates the record based on receiving a terminal error code and retries the communication based on receiving a retry error code.
Some embodiments provide systems to control digital communications comprising: a transceiver; a processing resource; and a medium storing instructions executed to cause the processing resource to: during a current session, determine acquisition execution intent prediction features; trigger a query to a first trained model; identify a first inferred acquisition execution type based on the acquisition execution intent prediction features; repeatedly evaluate according to a predefined interval whether to trigger a refresh of the first inferred acquisition execution type; trigger a first refresh query; identify, using the first trained model, a second inferred acquisition execution type; identify a set of one or more items corresponding to the in-session search; filter the set to a sub-set of items that satisfy the search and comply with the second inferred acquisition execution type; and control the data communications transceiver to transmit response data in controlling the remote client computing device to render content.
H04L 65/1089 - Procédures en session en ajoutant des médiasProcédures en session en supprimant des médias
G06F 12/0813 - Systèmes de mémoire cache multi-utilisateurs, multiprocesseurs ou multitraitement avec configuration en réseau ou matrice
H04L 41/16 - Dispositions pour la maintenance, l’administration ou la gestion des réseaux de commutation de données, p. ex. des réseaux de commutation de paquets en utilisant l'apprentissage automatique ou l'intelligence artificielle
H04L 65/1096 - Fonctions supplémentaires, p. ex. renvoi d'appel ou mise en attente d'appel
H04L 65/80 - Dispositions, protocoles ou services dans les réseaux de communication de paquets de données pour prendre en charge les applications en temps réel en répondant à la qualité des services [QoS]
Examples provide image quality assessment using computer vision (CV) object detection and recognition with depth estimation. An image quality manager obtains image quality analysis data, including CV object recognition results and depth information for objects of interest. The image analysis data is analyzed to identify image quality issues present in the images. The type of image quality issues includes object detection type, depth type, and payload type issues, such as images with inconsistent distance from an object of interest, images in which the object of interest is either too close or too far away, images having excessive time gaps between images, payloads with too few images, payloads without object detections, etc. Image quality feedback identifying the type of image quality issues detected is generated and provided to users. The system uses image quality feedback to retrain the CV models. The feedback optionally includes suggested actions for resolving the detected issues.
G06T 7/55 - Récupération de la profondeur ou de la forme à partir de plusieurs images
G06V 10/25 - Détermination d’une région d’intérêt [ROI] ou d’un volume d’intérêt [VOI]
G06V 10/26 - Segmentation de formes dans le champ d’imageDécoupage ou fusion d’éléments d’image visant à établir la région de motif, p. ex. techniques de regroupementDétection d’occlusion
G06V 10/764 - Dispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique utilisant la classification, p. ex. des objets vidéo
G06V 10/776 - ValidationÉvaluation des performances
G06V 10/94 - Architectures logicielles ou matérielles spécialement adaptées à la compréhension d’images ou de vidéos
27.
SYSTEMS AND METHODS FOR PREDICTING FULFILLMENT TIMES AND UPDATING USER REQUESTS
Example implementations relate to predicting fulfillment times for user requests. In an example, a user request including item data, a fulfillment location, and a provisional fulfillment time is received by a system. The system requests, via a fulfillment predictor, fulfillment data for the fulfillment location. The fulfillment data including at least a fulfillment order for the user request and ongoing fulfillment orders for the user request. The system, in response to receiving the fulfillment data, determines, by the fulfillment predictor, a predicted fulfillment time for the user request. The system determines whether a difference between the provisional fulfillment time and the predicted fulfillment time is within a fulfillment time threshold and, in accordance with a determination that the difference between the provisional fulfillment time and the predicted fulfillment time is outside the fulfillment time threshold, transmits the predicted fulfillment time to at least one computing device.
G06Q 10/04 - Prévision ou optimisation spécialement adaptées à des fins administratives ou de gestion, p. ex. programmation linéaire ou "problème d’optimisation des stocks"
G06Q 10/087 - Gestion d’inventaires ou de stocks, p. ex. exécution des commandes, approvisionnement ou régularisation par rapport aux commandes
05 - Produits pharmaceutiques, vétérinaires et hygièniques
29 - Viande, produits laitiers et aliments préparés ou conservés
30 - Aliments de base, thé, café, pâtisseries et confiseries
31 - Produits agricoles; animaux vivants
32 - Bières; boissons non alcoolisées
Produits et services
(1) Food for babies; Pharmaceuticals, medical and veterinary preparations; sanitary preparations for medical purposes; dietetic food and substances adapted for medical or veterinary purposes; dietary supplements for human beings and animals; adhesive plasters, materials for dressings; material for filling teeth, dental wax; disinfectants; preparations for destroying vermin; fungicides, herbicides.
(2) Meat, fish, poultry and game; meat extracts; preserved, frozen, dried and cooked fruits and vegetables; jellies, jams, compotes; eggs; milk, cheese, butter, yogurt and other milk products; oils and fats for food; nut butter, potato chips; Meat extracts for culinary purposes; preserved, frozen, dried and cooked fruits, vegetables and seaweeds.
(3) Coffee, tea, cocoa and substitutes therefor; rice, pasta and noodles; tapioca and sago; flour and preparations made from cereals; bread, pastries and confectionery; chocolate; ice cream, sorbets and other edible ices; sugar, honey, treacle; yeast, baking-powder; salt, seasonings, spices, preserved herbs; vinegar, sauces and other condiments; ice (frozen water).
(4) Raw nuts; Raw and unprocessed agricultural, aquacultural, horticultural and forestry products; raw and unprocessed grains and seeds; fresh fruits and vegetables, fresh herbs; natural plants and flowers; bulbs, seedlings and seeds for planting; live animals; foodstuffs and beverages for animals; malt.
(5) Beers; non-alcoholic beverages; mineral and aerated waters; fruit beverages and fruit juices; syrups and other preparations for making non-alcoholic beverages; sparkling water; expressly excluding vegetable-based food drinking products.
21 - Ustensiles, récipients, matériaux pour le ménage; verre; porcelaine; faience
Produits et services
Lunch boxes; Bento boxes; Water bottles sold empty; Plastic water bottles sold empty; Insulated containers for food or beverages; Reusable stainless steel water bottles sold empty; Tumblers for use as drinking glasses
41 - Éducation, divertissements, activités sportives et culturelles
Produits et services
Educational services, namely, developing, arranging, and conducting educational conferences and programs and providing courses of instruction in the field of information security; Educational services, namely, providing educational speakers in the field of information security; Providing an in-person educational forum in the field of information security
Examples may be related to playlist data for media devices. An example may involve receiving a network request including at least one campaign reservation request; generating, based on the network request, playlist data associated with a time period and a physical store; and determining one or more media devices that are located at the physical store and are available to display the playlist during the time period. The playlist data may be transmitted to the one or more media devices for displaying the playlist on the one or more media devices within the time period.
In some embodiments, systems and methods are provided herein useful to generating item recommendations. In some embodiments, a system for item recommendation may include a memory storing an item recommendation model trained using dense representations for items, the dense representations generated based on an item interaction graph including nodes corresponding to items for sale in a retail facility, edges connecting the nodes, and edge weights for the edges, wherein the edge weights are related to item selection sequence of the items, and a control circuit coupled to the memory, the control circuit configured to receive an indication of an item selection by a user in the retail facility, generate a recommendation for a suggested next item based on using the item selection as an input of the item recommendation model, and cause the one or more suggested next best items to be presented to the user for potential selection.
Example implementations relate to generating an interaction analysis that includes receiving interaction data for an interface element included in a user interface during a time period coinciding with an interaction campaign. If an initial automated analysis of the interaction data does not meet a first predetermined threshold, a set of time series features are generated for at least a portion of the time period. Synthetic interaction data for the interface element are generated during at least the portion of the time period. The synthetic interaction data represents interactions with the interface element independent of the interaction campaign and is generated by a time series model that receives the set of time series features. A difference metric for the interaction data and the synthetic interaction data is determined and if the difference metric is above a second predetermined threshold, the difference metric is stored in a database.
17 - Produits en caoutchouc ou en matières plastiques; matières à calfeutrer et à isoler
24 - Tissus et produits textiles
Produits et services
Packaging materials made of corrugated cardboard; Corrugated cardboard for packaging; Cushioning or padding made of paper for packing purposes; Plastic bubble packs for wrapping or packaging; Plastic bags for packaging; Adhesive packing tape for stationery or household use; Moving boxes of cardboard; Plastic wrap Molded foam for packing; Low-density polyurethane foam for packing Unfitted furniture covers not of paper
17 - Produits en caoutchouc ou en matières plastiques; matières à calfeutrer et à isoler
24 - Tissus et produits textiles
Produits et services
Packaging materials made of corrugated cardboard; Corrugated cardboard for packaging; Cushioning or padding made of paper for packing purposes; Plastic bubble packs for wrapping or packaging; Plastic bags for packaging; Adhesive packing tape for stationery or household use; Moving boxes of cardboard; Plastic wrap Molded foam for packing; Low-density polyurethane foam for packing Unfitted furniture covers not of paper
Examples provide an archway truss for supporting sensor devices associated with a lane passing through the archway. The sensor de generate sensor data associated with objects moving through the archway. The archway truss includes vertical support members attached to a horizontal top member. A central support member is reinforced to protect against cart impacts. One or more cameras are attached to the archway truss and positioned to capture images of objects in carts moving through the archway truss. Radio frequency identification (RFID) tag readers and other sensor devices are removably attached to the archway truss to gather item identification data for objects in the carts. Pairs of wing barriers are provided for each vertical support to block the field of view of the cameras from objects outside the lanes formed by the archway truss. An exterior covering provides padding to protect users contacting the archway truss.
Example implementations relate to detecting a terminated entity in a network environment. A network activity dataset including data representative of network activity within a network environment and a plurality of data records is received. Each data record in the plurality of data records includes a set of attributes. A graph that links systems having a first role in the data representative of network activity and a subset of the plurality of data records is generated. Feature information from the set of attributes for one or more data records in the subset of the plurality of data records in the graph is aggregated. A machine learning model is trained based on the aggregated feature information derived from the graph. Using the trained model, a determination representing a likelihood that a respective system having the first role in the data representative of network activity is linked to the terminated entity is generated.
Example implementations relate to order prediction and automated cart creation. In an example, basket features, item features, and order features associated with a profile are inputted into a predictive model and daily order predictions are outputted by the predictive model. The daily order predictions are inputted into a grouping algorithm to output an order predictability. Profiles are segmented into respective order predictability cohorts of a plurality of order predictability cohorts based on respective order predictabilities. A cart may be automatically created for at least one daily order prediction associated with at least one profile when a respective order predictability cohort belongs to a predetermined order predictability cohort.
Example implementations relate to anomaly detection in a network environment. In an example, a similarity score for one or more attributes between a target user and a candidate user is calculated based on n-grams generated from the one or more attributes. Link data linking the target user to the first candidate user for the first attribute is generated if the similarity score between the target user and the first candidate user is greater than a first threshold. A machine learning model that identifies a likelihood whether the target user is linked to a terminated entity based on the one or more attributes is trained using the link data. The machine learning model applies respective weights to each of the one or more attributes. The respective weights associated with the one or more attributes is updated based on feedback data associated with changes in operating permissions within a predetermined time period.
G06N 3/043 - Architecture, p. ex. topologie d'interconnexion fondée sur la logique floue, l’appartenance floue ou l’inférence floue, p. ex. systèmes d’inférence neuro-floue adaptatifs [ANFIS]
Clothing, namely, t-shirts, socks, underwear On-line retail store services featuring a wide variety of consumer goods; Retail store services featuring a wide variety of consumer goods of others; Retail store services featuring apparel.; Retail store services featuring a wide variety of consumer goods; Retail store services featuring a wide variety of consumer goods of others; Retail store services featuring apparel.
Example implementations relate to input mitigation for automated systems. In an example, a user input for a targeted model is received an anomaly score of the user input is determined. The anomaly score is representative of a similarity to expected inputs for the target model. In response to determining the anomaly score equal to or greater than a predetermined threshold, the user input is prevented from being provided to the target model and a mitigation process is implemented based on the user input. In response to determining the anomaly score is less than the predetermined threshold, the user input is provided to the target model and a responsive output based on the user input is generated by the target model.
Systems and methods for topic modeling of conversational data are disclosed. In an example, text data is received and is related to one or more topics. Text data is provided to a first machine learning model perform text preparation tasks. A summarization of the text data is generated using a second machine learning model. Clusters of the summarization are determined using a third machine learning model. The one or more topics are identified using a fourth machine learning model. Subtopics of each of the identified one or more topics are identified by recursively using the fourth machine learning mode. An output of the summarization, the identified one or more topics, and the identified one or more subtopics of each of the identified one or more topics is generated. The generated output is displayed to a user and feedback on the output is received.
A method including determining respective training data for each of query-type-specific answer retrieval modules. The method further can include training each of the query-type-specific answer retrieval modules. The method additionally can include determining a query type of a query from a user device for a user. The method also can include determining, in real-time and based at least in part on the query type, an answer retrieval module from the query-type-specific answer retrieval modules, as trained. Moreover, the method can include determining, in real-time by the answer retrieval module, one or more answers for the query. Then, the method can include ranking, in real-time, the one or more answers based on a user profile of the user. Finally, the method can include transmitting, via a computer network and to the user device, at least one of the one or more answers, as ranked. Other embodiments are disclosed.
Examples provide an autonomous application programming interface (API) agent for executing multiple APIs simultaneously in parallel or in sequence. The API agent performs a vector database similarity search using embeddings representing a user query and a plurality of candidate APIs to identify one or more relevant APIs for responding to the user query. The API agent utilizes a large language model (LLM) API orchestrator for performing dynamic slot filling to replace missing API parameters required for calling one or more of the relevant APIs. The API agent executes the query on the relevant APIs to obtain response data from the relevant APIs. The response data is filtered and combined into a single query response which is provided to a user in response to the query.
Examples provide a future inventory (FI) ordering system that enables orders of temporarily out-of-stock (OOS) items that are not currently on-hand at a fulfillment center (FC). A temporarily unavailable OOS item is made available for purchase with an extended estimated date of delivery (EEDD) as a future delivery (FD) item. A status indicator can be provided to distinguish FD items from currently in-stock items via a user interface (UI) device. Machine learning models are used to predict transit time, dwell time, and/or receiving time for the FD item using lane-specific data, dynamic extrinsic data, and other item-related data. The EEDD is predicted using the predicted transit time. A delivery notification including the EEDD for the FD item to the user via the UI device. Unloading of trailers containing FD items is prioritized at the FC to ensure timely delivery of FD items within the predicted EEDD.
Example implementations related to automated testing of processes are disclosed. In an example, historical interaction data is annotated to generate annotated interaction data. A synthetic user definition including one or more machine-interpretable instructions for tuning a language model is generated based on the annotated interaction data and a simulation request for implementing a test of at least one trained process is received. User interactions with the at least one trained process by a synthetic user are simulated based on the synthetic user definition. The simulated user interaction generates an interaction output. A determination is made whether at least one goal of the simulated user interaction was met and the interaction output is labeled based on the determination whether the at least one goal was met.
Examples provide a multi-object cluster detection model for identifying instances of items arranged in clusters of objects of interest using input images of a selected area generated by an image capture device. A cluster detection manager identifies instances of each different cluster of items within the image based on size, shape, and color of the objects as well as the proximity of the items to one another. Labeled image data is generated using the image. The labeled image data includes cluster indicators identifying unique clusters in the image. The labeled image data is used to determine the number of different clusters in a location. The system generates an alert if the detected number of clusters differs from an expected number of clusters for the given location. The system utilizes cluster detections to update item data, identify out-of-stock items, identify incorrectly placed items, validate automated item-to-location mapping, and verify item identifications.
G06V 10/762 - Dispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique utilisant le regroupement, p. ex. de visages similaires sur les réseaux sociaux
Example implementations relate to automated ticket classification and response generation. In an example, historical tickets, each including at least one text field, are received, and an embedding model is applied to generate at least one embedding for each historical ticket. The historical tickets are clustered in a plurality of clusters. For each cluster, keywords are extracted from the text fields of each historical ticket in the cluster and a cluster-specific feature matrix is generated. A ticket is received, and at least one feature embedding is generated for the ticket. A similar cluster is determined for the ticket based on similarity between the at least one feature embedding for the ticket and the at least one cluster-specific feature embedding for each cluster in the plurality of clusters. A generative model is applied to generate one or more response steps based on historical response steps and to generate a response output.
Examples related to predicting user behaviors are disclosed. An example may involve: obtaining current behavior data of a user within a current user session; obtaining historical behavior data of the user during a past time period; determining, using at least one natural language model, context data that is relevant for predicting future behavior data of the user, wherein the context data is determined based on the current behavior data and the historical behavior data; generating, using a prediction model, a ranked list of elements related to the future behavior data of the user based on the context data; and transmitting the ranked list of elements to a computing device associated with the current user session.
A computer implemented method including determining an expected click-through-rate (CTR) of a query-item pair with a first machine learning model by using content-based features. The computer implemented method can also include, determining a click engagement (CE) feature by determining a Bayesian inference based on the expected CTR and a historical CTR for the query-item pair. The computer implemented method can further include, determining a rerank score of the query-item pair with a second machine learning model by using the content-based features and the CE feature. The computer-implemented method can additionally include reranking the query-item pair based in part on the rerank score and the expected CTR. Other embodiments are described.
Example implementations relate to automatically generating and updating functional tests for software applications. A first version of requirement information for a software application is received and segmented into respective portions. An input to a trained model is generated based on the respective portions. One or more functional tests that determine whether the software application satisfies the first version of the requirement information are generated. Embeddings associated with the one or more functional tests are generated. In response to receiving an updated version of the requirement information for the software application: one or more semantic differences between the versions of the requirement information are identified, a modified input to the trained model is generated based on the semantic differences and the embeddings, one or more updated functional tests are generated to determine whether the software application satisfies the updated version of the requirement information.
Examples provide a system for generating image-based training data using progressive data curation. An anchor image of a selected item and historical receipts including the selected item generated during a dynamic receipt retrieval time period are obtained. Images of the carts including the selected item paired with the receipts are analyzed and cropped to isolate the selected item from each cart image. An embedding model generates embeddings representing the anchor image and the cropped images of the selected item. A similarity of the cropped image embeddings to the anchor image embedding is calculated using a similarity metric. The cropped image embeddings are ranked based on the calculated similarity to the anchor image. The images having the highest rank and greatest similarity to the anchor image are selected for inclusion in training data used to train computer vision models to detect and/or recognize the selected item in images of various objects.
G06V 10/774 - Génération d'ensembles de motifs de formationTraitement des caractéristiques d’images ou de vidéos dans les espaces de caractéristiquesDispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique utilisant l’intégration et la réduction de données, p. ex. analyse en composantes principales [PCA] ou analyse en composantes indépendantes [ ICA] ou cartes auto-organisatrices [SOM]Séparation aveugle de source méthodes de Bootstrap, p. ex. "bagging” ou “boosting”
A system is provided including a processor and a non-transitory computer-readable medium storing computing instructions that, when executed on the processor, cause the processor to perform operations including: obtaining a plurality of query-item pairs associated with a query, wherein the plurality of query-item pairs associated with the query are initially ranked at least partially based on user engagement metrics; obtaining, for each query-item pair of the plurality of query-item pairs associated with the query, first temporal-behavioral features, second temporal-behavioral features, and context-aware features, wherein the context-aware features include a vertical of the query; generating, using a Gradient Boosted Decision Tree (GBDT) model, respective query-item pair GBDT output values based on respective inputs for each query-item pair of the plurality of query-item pairs associated with the query, wherein the respective inputs for each query-item pair of the plurality of query-item pairs associated with the query are based on the first temporal-behavioral features, the second temporal-behavioral features, and the context-aware features; and re-ranking at least some of the plurality of query-item pairs associated with the query that are initially ranked based on the respective query-item pair GBDT output values.
Examples provide a multi-object cluster detection model for identifying instances of items arranged in clusters of objects of interest using input images of a selected area generated by an image capture device. A cluster detection manager identifies instances of each different cluster of items within the image based on size, shape, and color of the objects as well as the proximity of the items to one another.
Examples provide a multi-object cluster detection model for identifying instances of items arranged in clusters of objects of interest using input images of a selected area generated by an image capture device. A cluster detection manager identifies instances of each different cluster of items within the image based on size, shape, and color of the objects as well as the proximity of the items to one another.
Labeled image data is generated using the image. The labeled image data includes cluster indicators identifying unique clusters in the image. The labeled image data is used to determine the number of different clusters in a location. The system generates an alert if the detected number of clusters differs from an expected number of clusters for the given location. The system utilizes cluster detections to update item data, identify out-of-stock items, identify incorrectly placed items, validate automated item-to-location mapping, and verify item identifications.
G06V 10/762 - Dispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique utilisant le regroupement, p. ex. de visages similaires sur les réseaux sociaux
In some embodiments, apparatuses and methods are provided herein useful to processing captured images. In some embodiments, there is provided a system for processing captured images of objects at a product storage facility including a trained machine learning model stored in a memory; and a control circuit. The control circuit may obtain an image at the product storage facility; cluster objects depicted in the image that have same product identifiers into a corresponding group; determine coordinates of each bounding box of each clustered object in the corresponding group; determine a bounding box representative depth value of pixels inside the bounding box of each clustered object; determine an overall representative depth value of the corresponding group based on bounding box representative depth values of clustered objects; and exclude the clustered objects from identified objects in the image upon a determination that the overall representative depth value is greater than a threshold.
G06T 7/70 - Détermination de la position ou de l'orientation des objets ou des caméras
G06T 7/50 - Récupération de la profondeur ou de la forme
G06V 10/25 - Détermination d’une région d’intérêt [ROI] ou d’un volume d’intérêt [VOI]
G06V 10/762 - Dispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique utilisant le regroupement, p. ex. de visages similaires sur les réseaux sociaux
75.
COMPUTING PRODUCT DIMENSIONS USING TEXT-BASED FEATURES
A method can include computing, a dimension of a product using a machine learning model, a dimension of a product. The computing can include, transforming preprocessed data into numerical values using a vectorize layer. The computing can also include, condensing the numerical values using an embedding layer. The predicting can further include, processing the numerical values, as condensed, using a global average pooling layer. The predicting can also include, concatenating the numerical values, as processed into a layer. The predicting can further include processing the layer using a first dense layer. The predicting can additionally include, standardizing the layer, as processed, using a batch normalization layer. The predicting can also include, transforming the layer, as standardized, using an activation function. The predicting can additionally include, regularizing the layer, as transformed, using a dropout function. The predicting can further include, outputting the dimension of the product, with an output layer. Other embodiments are described.
Examples relate to managing node inventory levels are disclosed. An example may involve: obtaining historical data of items at a node in a network for a past time period; generating a supply chain lead time variability for an item at the node based on the historical data; generating a demand variability for the item at the node based on the historical data and the supply chain lead time variability; generating, based on the supply chain lead time variability and the demand variability, a recommended inventory level of the item in the node for a future time period; and transmitting the recommended inventory level to a computing device associated with the node for inventory arrangement in the future time period.
Example implementations relate to interface generation including interface elements representative of a co-located resource when a user device is located within a predetermined geofenced area and a corresponding user has historical interactions with the co-located resource. In an example, a request for an interface and location data are received from a user device. The location data corresponds to a location of the user device when the request for the interface was generated. In response to determining the location data is within a predetermined geofenced area, a resource use probability of a resource for the user device is generated using a resource affinity model based at least in part on historical usage of the resource. In response to determining the resource use probability is above a predetermined threshold, instructions are transmitted to the user device that modify the interface to include an interface element representative of a resource usage.
H04W 4/021 - Services concernant des domaines particuliers, p. ex. services de points d’intérêt, services sur place ou géorepères
H04L 41/22 - Dispositions pour la maintenance, l’administration ou la gestion des réseaux de commutation de données, p. ex. des réseaux de commutation de paquets comprenant des interfaces utilisateur graphiques spécialement adaptées [GUI]
Examples provide a system and method for dynamically filtering candidate item identifiers (IDs) from a pool of item IDs in real-time for automatic labeling of images for use as training data used to train computer vision (CV) models. Images of carts are paired with item receipts. Candidate item IDs are extracted from the receipts. Item recognition inference results generated by CV models are used to pair images of individual items with item IDs identifying the item in each item image. As each candidate item ID is assigned to an item image, the item ID is dynamically filtered. Any candidate item IDs remaining after filtering are assigned to any item images failing to pair with an item ID based on the infer results. The results are presented for review and status update via a user interface device for faster and more accurate auto-labeling of training data for CV models.
G06V 20/70 - Étiquetage du contenu de scène, p. ex. en tirant des représentations syntaxiques ou sémantiques
G06V 10/764 - Dispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique utilisant la classification, p. ex. des objets vidéo
Example implementations related to next state prediction and action selection are disclosed. In an example, an initial state is estimated based on a set of priors including at least one engagement opportunity. A diminishing effect of the at least one engagement opportunity is determined using a non-linear transformation and a set of likelihood estimates for the initial state is generated using a Bayesian steady state filter model that receives the initial state and the diminishing effect of the at least one engagement opportunity. A next state is predicted based on the initial state and the set of likelihood estimates.
(1) Retail and online department store services; retail and online drug store services; retail and online grocery store services; retail and online services provided by hypermarkets; operation of customer incentive, award and loyalty programs; consumer membership program services.
(1) Retail and online department store services; retail and online drug store services; retail and online grocery store services; retail and online services provided by hypermarkets; operation of customer incentive, award and loyalty programs; consumer membership program services.
(1) Retail and online department store services; retail and online drug store services; retail and online grocery store services; retail and online services provided by hypermarkets; operation of customer incentive, award and loyalty programs; consumer membership program services.
Wholesale and retail store services featuring a wide variety of consumer goods of others; Retail store services featuring a wide variety of consumer goods of others; On-line retail store services featuring a wide variety of consumer goods of others
09 - Appareils et instruments scientifiques et électriques
35 - Publicité; Affaires commerciales
36 - Services financiers, assurances et affaires immobilières
42 - Services scientifiques, technologiques et industriels, recherche et conception
Produits et services
Downloadable computer software featuring radio-frequency identification (RFID) technology, machine learning algorithms, and computer vision technology, namely, cameras and object detection and recognition models for analyzing objects in real time, for internal retail store use in recognition of physical objects in carts to confirm accuracy of consumer goods sales transactions in retail store environments; Electronic exit verification systems comprised primarily overhead cameras, RFID readers, handheld computing devices, and recorded operating software featuring machine learning models and computer vision technology namely, cameras and object detection and recognition modes for analyzing objects in real time, for internal retail store use in recognition of physical objects in carts to verify accuracy of consumer goods sales transactions in retail store environments Retail store exit monitoring services for business and accounting purposes featuring radio-frequency identification (RFID) technology, and machine learning algorithms, and computer vision technology, namely, cameras and object detection and recognition models for analyzing objects in real time, for internal retail store use in verification of consumer goods sales transactions in retail store environments Consumer goods payment verification services by means of radio-frequency identification (RFID) technology, machine learning algorithms, and computer vision technology, namely, cameras and object detection and recognition models for analyzing objects in real time, for internal retail store use in retail store environments; monitoring of financial processing by means of RFID technology, machine learning algorithms, and computer vision technology, namely, cameras and object detection and recognition models for analyzing objects in real time, for internal retail store use in retail store environments Monitoring and verification of consumer goods sales transactions by means of radio-frequency identification (RFID) technology, machine learning algorithms, and computer vision technology namely, cameras and object detection and recognition models for analyzing objects in real time, for internal retail store use in retail environments
86.
ACTIVE LEARNING FOR DETECTION LABELING VIA FOUNDATION MODELS
Examples provide active learning for effective computer vision (CV) item detection labeling using foundation models to generate updated training data for retraining CV item detection models. Raw image data of shopping carts in a retail facility are analyzed by a pretrained CV item detection model to identify items in the carts. The detected items are labeled and enclosed in bounding boxes. A set of foundation models mask the detected items in the cart images. Predicted labels for the undetected and unmasked items in the cart images are generated. Predicted bounding boxes enclosing the unmasked items undetected by the CV item detection model are generated. The predicted bounding boxes and predicted labels are merged with the detected items bounding boxes and labels to generate updated training data for dynamically retaining the CV item detection model to detect future occurrences of the undetected items in cart images with greater accuracy and efficiency.
G06V 10/774 - Génération d'ensembles de motifs de formationTraitement des caractéristiques d’images ou de vidéos dans les espaces de caractéristiquesDispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique utilisant l’intégration et la réduction de données, p. ex. analyse en composantes principales [PCA] ou analyse en composantes indépendantes [ ICA] ou cartes auto-organisatrices [SOM]Séparation aveugle de source méthodes de Bootstrap, p. ex. "bagging” ou “boosting”
G06V 20/70 - Étiquetage du contenu de scène, p. ex. en tirant des représentations syntaxiques ou sémantiques
Examples provide a system and method for automatically tagging items and forming item-level user cohorts using a generative artificial intelligence (GenAI) model. A GenAI model generates item-level persona labels for a subset of untagged items sampled from a category of items. An item-level persona reflects user behaviors and/or preferences associated with specific items and types of items. The labeled item data, including the persona identification (ID) label, is used to train a deep neural net (DNN) labeling model to label untagged items with persona IDs for item-level personas. Customized item recommendations for each tagged item is mapped to a user cohort and used to generate item recommendations customized at the item-level. The DNN labeling model is periodically tested, evaluated, and retrained using sample tagged item data from the GenAI model to reduce the DNN model error rate and improve accuracy of the DNN model persona predictions.
Examples provide a system for enhanced pallet label checking using computer vision and optical character recognition for faster and more efficient resolution of pallet label exceptions. The system includes a pallet manager component that obtains images of pallets from one or more image capture devices. Computer vision and machine learning is utilized to identify pallet labels on pallets which are missing or damaged such that the pallet labels are at least partially unreadable. An initial pallet label exception is created. The exceptions are assigned scores indicating a degree of confidence that the exceptions are accurate and require attention to resolve the issues associated with each label. The exceptions having high confidence scores are enhanced with customized label check instructions and real time images of the pallets. The enhanced pallet label exceptions assist users in locating pallets and resolving issues associated with pallet labels with greater speed and accuracy.
A coupler for coupling a vacuum source to a rotatable tool coupled to an end of a rotatable robotic arm for moving at least one object includes a first portion coupled to both the rotatable tool and the rotatable robotic arm. The first portion and the rotatable tool are configured to rotate in response to rotation of the end of the robotic arm. The first portion includes a first conduit passing therethrough and being in communication with the vacuum source. The coupler further includes a second portion that does not rotate during rotation of the rotatable robotic arm and the rotatable tool. The coupler further includes a second conduit configured to couple to a first end of a flexible tube having a second end thereof coupled to the vacuum source. The second conduit does not rotate during the rotation of the rotatable robotic arm and the rotatable tool.
B25J 15/06 - Têtes de préhension avec moyens de retenue magnétiques ou fonctionnant par succion
B25J 19/00 - Accessoires adaptés aux manipulateurs, p. ex. pour contrôler, pour observerDispositifs de sécurité combinés avec les manipulateurs ou spécialement conçus pour être utilisés en association avec ces manipulateurs
Example implementations relate to automated modification of interface pages to include relevant link information. In an example, an anchor page and a plurality of candidate pages are obtained and an embedding is generated for each page by applying a page embedding model to a representative keyword construction extracted from each of the at least one anchor page or a candidate page of the plurality of candidate pages. A set of similar candidate pages is selected using a comparison of the embeddings and a relevance score is determined for each candidate page. The relevance score represents a relevance of a candidate page to the anchor page. A diversification score is generated for each candidate page and a linked set of candidate pages is selected based on the relevance scores and diversification scores. The anchor page is modified to include a link to each page in the linked set.
