Abstract: Examples described herein include systems and methods for providing a workflow on a user device. A user device can receive image data from a camera of the user device. The device can perform object recognition to recognize an object from the image data, as well as classification of the object. The user device can match the object classification with a backend system. The example method can also include launching a workflow form associated with the matching backend system. The user device can launch a form having these fields available, either within the workflow application or within a dedicated application associated with the relevant backend system. The user device can pre-fill one or more information fields in the workflow form based on the object classification, providing the user with a head start on submitting the form to the appropriate system.

Abstract: Aerial imagery may be captured from an unmanned aircraft or other aerial vehicle. The use of aerial vehicle imagery provides greater control over distance from target and time of image capture, and reduces or eliminates imagery interference caused by clouds or other obstacles. Images captured by the aerial vehicle may be analyzed to provide various agricultural information, such as vegetative health, plant counts, population counts, plant presence estimation, weed presence, disease presence, chemical damage, wind damage, standing water presence, nutrient deficiency, or other agricultural or non-agricultural information. Georeference-based mosaicking may be used to process and combine raw image files into a direct georeferenced mosaic image.

Abstract: The method for determining property feature segmentation includes: receiving a region image for a region; determining parcel data for the region; determining a final segmentation output based on the region image and parcel data using a trained segmentation module; optionally generating training data; and training a segmentation module using the training data S500.

Abstract: Disclosed are methods, devices, and computer-readable media for operating paired or grouped drone devices. In one embodiment, a method is disclosed comprising capturing a first image by a camera installed on a first drone device; processing the first image using an artificial intelligence (AI) engine embedded in the first drone device, the processing comprising generating a first inference output; transmitting the first inference output to a second drone device; receiving a second inference output from the second drone device, the second inference output associated with a second image captured by the second drone device; and transmitting the first image to a processor based on the first inference output and second interference output.

Abstract: Disclosed is an infrared thermal image classification and hot spot positioning method of a photovoltaic panel, comprising following steps: constructing a photovoltaic panel infrared thermal image data set, preprocessing the image data set, and dividing into a training set and a testing set according to a preset proportion; training an auxiliary generating countermeasure network based on the training set to obtain a trained generator and a trained discriminator, training an encoder by combining the image data in the training set with the generator, and fixing the parameters of a trained encoder; inputting the image data in the test set into a trained discriminator to obtain a photovoltaic image classification result; inputting images in the test set into the trained encoder, and then inputting the images into the generator for a reconstruction; comparing pixels of the input images with the reconstructed infrared thermal images, and positioning hot spots.

Abstract: An error detection system for notifying maintenance personnel of errors in a bowling lane operation may include a camera, artificial light source (optional), a computer that sends an output signal. The output signal may be a signal to provide notification to maintenance personnel to fix the error or an electrical signal to the pinsetter or ball return unit to manipulate operation of the same to prevent further damage. In this manner, interruptions and delays to bowlers will be minimized because maintenance personnel is immediately notified of any errors and can resolve the errors before the bowler realizes the error.

Abstract: Disclosed herein include embodiments of a system, a device, and a method for sorting a plurality cells of a sample. A plurality of raw images comprising pixels of complex values in a frequency space can be generated from a plurality of channels of fluorescence intensity data of fluorescence emissions of fluorophores, the fluorescence emissions being elicited by fluorescence imaging using radiofrequency-multiplexed excitation in a temporal space. Spectral unmixing can be performed on the raw images prior to a sorting decision being made.

Abstract: A medical testing apparatus for a user to help to authenticate and track the process of a test kit that is received into the medical test apparatus, the test kit includes components of an extended nasal/oral swab, a buffer tube, a reagent disposed within the buffer tube, a nozzle cap, and a test card, the kit is for testing for a presence or a non-presence of an infection that utilizes an internet connected electronic device with a camera for communication to third parties. The apparatus includes a test stand that is configured to partially receive the electronic device and to partially receive the test kit in a positional alignment to for the camera to record an unique identification code for each test kit component, and for the camera user facial recognition, further the test stand camera records infection status.

Abstract: A medical image analysis method includes: reading an original medical image; performing image classification and object detection on the original medical image to generate a first classification result and a plurality of object detection results by a plurality of complementary artificial intelligence (AI) models; performing object feature integration and transformation on a first detection result and a second detection result among the object detection results to generate a transformation result by a features integration and transformation module; and performing machine learning on the first classification result and the transformation result to generate an image interpretation result by a machine learning module and display the image interpretation result.

Abstract: Computer-implemented methods and systems are provided for generating encoded representations for multiple magnifications of one or more query images. An example method for involves operating a processor to obtain a query image and identify a set of anchor points within the query image. The processor is operable to generate a plurality of sub-images for a plurality of magnification levels for each anchor point. Each sub-image includes the anchor point and corresponds to a magnification level of the plurality of magnification levels. The processor is operable to, for each magnification level, apply an artificial neural network model to a group of sub-images having that magnification level to extract a feature vector representative of image characteristics of the query image at that magnification level; and to generate an encoded representation for multiple magnifications of the query image based on the feature vectors extracted for the plurality of magnification levels.

Abstract: A system includes a computing device that receives a query thyroid image, where the query thyroid image is an ultrasound image of a thyroid comprising a thyroid nodule of interest. The computing device processes the query thyroid nodule image using a machine learning model to identify at least one labelled thyroid image from a plurality of labelled thyroid images that is similar to the query thyroid nodule image. The plurality of labelled thyroid images are used as training data to generate the machine learning model. The at least one labelled thyroid image has labels associated therewith and comprises an ultrasound image of a thyroid nodule that has a confirmed diagnosis. The computing device generates an output report based on the labels associated with the at least one labelled thyroid image, where the output report indicates whether the thyroid nodule of interest resembles a malignant thyroid nodule or benign thyroid nodule.

Abstract: The present invention relates to a method and system that automatically finds, segments and measures lesions in medical images following the Response Evaluation Criteria In Solid Tumours (RECIST) protocol. More particularly, the present invention produces an augmented version of an input computed tomography (CT) scan with an added image mask for the segmentations, 3D volumetric masks and models, measurements in 2D and 3D and statistical change analyses across scans taken at different time points. According to a first aspect, there is provided a method for determining volumetric properties of one or more lesions in medical images comprising the following steps: receiving image data; determining one or more locations of one or more lesions in the image data; creating an image segmentation (i.e. mask or contour) comprising the determined one or more locations of the one or more lesions in the image data and using the image segmentation to determine a volumetric property of the lesion.

Abstract: Systems and methods for detecting image anomalies include extracting one or more detected images from a submission file received from at least one computing device and generating an image identification (ID) for each of the one or more images. One or more image quality indices are determined for the submission file based on at least one of predetermined image features, an image type of the one or more images, and submission file attributes, and one or more image anomalies associated with the one or more images of the submission file are detected based on at least one of the image ID and the one or more image quality indices.

Abstract: The method for classifying ambiguous objects, including: determining initial labels for an image set; determining N training sets from the initially-labelled image set; training M annotation models using the N training sets; determining secondary labels for each image of the image set using the M trained annotation models; and determining final labels for the image set based on the secondary labels. The method can optionally include training a runtime model using images from the image set labeled with the final labels; and optionally using the runtime model.

Abstract: Embodiments for implementing enhanced ensemble model diversity and learning by a processor. One or more data sets may be created by combining one or more clusters of data points of a minority class with selected data points of a majority class. One or more ensemble models may be created from the one or more data sets using a supervised machine learning operation. An occurrence of an event may be predicted using the one or more ensemble models.

Abstract: In accordance with the present disclosure, deep-learning techniques are employed to find anomalies corresponding to bleed events. By way of example, a deep convolutional neural network or combination of such networks may be trained to determine the location of a bleed event, such as an internal bleed event, based on ultrasound data acquired at one or more locations on a patient anatomy. Such a technique may be useful in non-clinical settings.

Abstract: A method of selecting or generating a design for a component of a patient interface device, the method including receiving a scan of at least a portion of a patient's face, receiving preference information about the patient, and generating a custom design or selecting a preexisting design for the component of the patient interface device based on the scan of at least a portion of the patient's face and the preference information, wherein a size or shape of the generated custom design for the component or selected preexisting design for the component is based on the preference information.

Abstract: There is provided a method, comprising: providing a training dataset including, medical images and corresponding text based reports, and concurrently training a natural language processing (NLP) machine learning (ML) model for generating a NLP category for a target text based report and a visual ML model for generating a visual finding for a target image, by: training the NLP ML model using the text based reports of the training dataset and a ground truth comprising the visual finding generated by the visual ML model in response to an input of the images corresponding to the text based reports of the training dataset, and training the visual ML model using the images of the training dataset and a ground truth comprising the NLP category generated by the NLP ML model in response to an input of the text based reports corresponding to the images of the training dataset.

Abstract: A method for characterizing a pose estimation system includes: receiving, from a pose estimation system, first poses of an arrangement of objects in a first scene; receiving, from the pose estimation system, second poses of the arrangement of objects in a second scene, the second scene being a rigid transformation of the arrangement of objects of the first scene with respect to the pose estimation system; computing a coarse scene transformation between the first scene and the second scene; matching corresponding poses between the first poses and the second poses; computing a refined scene transformation between the first scene and the second scene based on coarse scene transformation, the first poses, and the second poses; transforming the first poses based on the refined scene transformation to compute transformed first poses; and computing an average rotation error and an average translation error of the pose estimation system based on differences between the transformed first poses and the second poses.

Abstract: A license plate detection and recognition system receives training data comprising images of license plates. The system prepares ground truth data from the training data based predefined parameters. The system trains a first machine learning algorithm based on the ground truth data to generate a license plate detection model. The license plate detection model is configured to detect one or more regions in the images. The one or more regions contains a candidate for a license plate. The LPDR system generates a bounding box for each region. The LPDR system trains a second machine learning algorithm based on the ground truth data and the license plate detection model to generate a license plate recognition model. The license plate recognition model generates a sequence of alphanumeric characters with a level of recognition confidence for the sequence.