A method for identifying a product removed from a cabinet includes detecting visual characteristics of the product removed from the cabinet by a camera within the cabinet. The method may include detecting an identifier on the product removed from the cabinet by an identifier sensor, and comparing the visual characteristics and the identifier to a database of product information. The product removed from the cabinet may be identified based on the comparison of the visual characteristics and the identifier to the database of product information.

When a software update is provided to a device that implements a facial recognition authentication process, a new authentication algorithm to operate the facial recognition authentication process may be included as part of software update. For a period of time, the new authentication algorithm may operate a “virtual” facial recognition authentication process alongside operation of the existing facial recognition authentication process using the existing (e.g., earlier version) authentication algorithm. The performance of the new authentication algorithm in providing facial recognition authentication (as assessed by the “virtual” process) may be compared to the performance of the existing authentication algorithm in providing facial recognition authentication during the period of time. When the performance of the new authentication algorithm is determined to have a satisfactory performance, operation of the actual facial recognition authentication process on the device may be switched to the new authentication algorithm.

A method for authenticating a food package holding a food product can include receiving sample image data depicting the food package, identifying a sample print feature sub-set of the food package, identifying a sample geometric feature sub-set of the food package, generating a sample feature set based on the sample print feature sub-set and the sample geometric feature sub-set, receiving a reference feature set, wherein the reference feature set is generated based on a reference print feature sub-set and a reference geometric feature sub-set identified in reference image data, wherein the reference image data is captured in a food packaging system arranged to produce food packages, comparing the sample feature set and the reference feature set, and in case of match, providing an indication that the food package is authentic.

A method and a device for determining a presence of tumor are provided. The method includes receiving a medical image associated with a patient. The medical image includes a region of interest associated with the patient. The method includes identifying one or more blood vessels associated with the region of interest in the medical image. The method includes determining a set of characteristics associated with the one or more blood vessels using a trained machine learning model. The method also includes determining whether the one or more blood vessels are feeder vessels associated with the tumor based on the set of characteristics associated with the one or more blood vessels. The method includes detecting a tumor region in the region of interest based on the feeder vessels, when the one or more blood vessels are the feeder vessels associated with the tumor.

Broadly speaking, the disclosure generally relates to relates to a computer-implemented methods and systems for scene graph generation, and in particular for training a machine learning, ML, model to generate a scene graph. The method includes inputting training a training image into a machine learning model, outputting a predicted label for at least two objects in the training image and a predicted label for a relationship between the at least two objects. The training method includes calculating a loss, which takes into account both a supervised loss calculated by comparing the predicted labels to the actual labels for the training image, and a logic-based loss calculated by comparing the predicted labels to stored integrity constraints comprising common-sense knowledge. Advantageously, this means that the performance of the model is improved without increasing processing at inference-time.

Methods, systems, and apparatuses are described for receiving image data and generating, based on the image data a set of instructions. The instructions may be configured to describe a method of weaving string around a loom so as to generate a string art representation of the received image.

A method includes receiving, at an artificial intelligence (AI) accelerator of a computing system, image data of one or more objects from an image sensor and performing an AI operation on the image data at the AI accelerator of the computing system using an AI model. The method further includes determining a custom cleaning cycle at the AI accelerator in response to performing the AI operation.

Methods, systems, articles of manufacture, and apparatus to extract shape features based on a structural angle template are disclosed. An example apparatus includes a template generator to generate a template based on an input image and calculate a template value based on values in the template; a bit slicer to calculate an OR bit slice and an AND bit slice based on the input image, combine the OR bit slice with the AND bit slice to generate a fused image, group a plurality of pixels of the fused image to generate a pixel window, each pixel of the pixel window including a pixel value, and calculate a window value based on the pixel values of the pixel window; and a comparator to compare the template value with the window value and store the pixel window in response to determining the window value satisfies a similarity threshold with the template value.

Systems, methods, and computer readable media for performing task assignment, completion, and management within a crowdsourced surveillance platform. A remote server may identify targets for image capture and may assign capture tasks to users based on travel plans of the user. Users may be assigned task to capture image of target locations lying along a travel path. The remote server may aggregate data related to the captured images and use it to update a map and log changes to the target location over time.

In an approach for determining the date of planting of a crop growing in an agricultural field, a processor receives an aerial image of one or more agricultural fields in a pre-determined geographical region. A processor selects a plurality of points from the aerial image. A processor calculates a Vegetation Index of one or more crops growing at the plurality of points selected. A processor compares the Vegetation Index calculated for the one or more crops growing at the plurality of points selected to the Vegetation Index known for a plurality of historical reference signatures. A processor generates an actual signature. A processor cross-correlates the actual signature against the plurality of historical reference signatures to measure a degree of similarity. A processor identifies the one or more crops growing in the one or more agricultural fields in the pre-determined geographical region from the cross-correlation.