Abstract: Embodiments of a kernel-level grain monitoring system include a grain camera positioned to capture bulk grain sample images of a currently-harvested grain taken into and processed by a combine harvester, a moisture sensor, and a display device. A controller architecture is coupled to the grain camera, to the moisture sensor, and to the display device. The controller architecture is configured to: (i) analyze the bulk grain sample images, as received from the grain camera, to determine an average per kernel (APK) volume representing an estimated volume of a single average kernel of the currently-harvested grain; (ii) repeatedly calculate one or more topline harvesting parameters based, at least in part, on the determined APK volume and the moisture sensor data; and (iii) selectively present the topline harvesting parameters on the display device for viewing by an operator of the combine harvester.

Abstract: A breast ultrasound diagnosis method using weakly supervised deep-learning artificial intelligence comprises: an ultrasound image preprocessing step of generating input data including only an image region necessary for learning, by deleting personal information about a patient from a breast ultrasound image; a deep-learning step of receiving the input data, obtaining a feature map from the received input data by using a convolutional neural network (CNN) and global average pooling (GAP), and carrying out re-learning; and a differential diagnosis step of determining the input data as one of normal, benign, and malignant by using the GAP, and when the input data is determined to be malignant, calculating a probability of malignancy (POM) indicating accuracy of the determination.

Abstract: A method and system is disclosed for acquiring image data of a subject. The image data can be collected with an imaging system having a detector able to move relative to the subject. A contrast agent can be injected into the subject and image data can be acquired with the contrast agent in various phases of the subject. A volumetric model of multiple phases can be reconstructed selected reconstruction techniques.

Abstract: A method for determining a position of corners of polygonal electrode sheets in at least one corner region of an electrode assembly stack, comprising the steps: 3D imaging of the corner region of the electrode assembly stack in an imaging region by means of an imaging process, so that a data set is created that includes 3D position information on the electrode sheets in the corner region of the electrode assembly stack relative to a carrier or to a marker arranged in the imaging region; and determining, from the data set, a first and a second edge profile of the edges framing the corner region of each electrode sheet. A position of the corner of the respective electrode sheet is determined on the basis of the edge profile. The electrode sheets and the edges thereof can be determined via a neural network system.

Abstract: Disclosed herein is an apparatus and method for counting repetitive movements based on Artificial Intelligence. The method may include generating standard movement information that includes a key pose extracted from a demonstration movement image stream based on human skeleton information, which is a set of pieces of positional information of human joints, and major joint information in the key pose and counting repetitive movements depending on whether a user movement matches the standard movement information based on human skeleton information of a user movement image stream.

Abstract: A computing device, method and computer program product are provided to generate training data for a machine learning system including training data representative of one or more edge scenes. In the context of a computing device, the computing device includes a simulator configured in accordance with a sampling algorithm to create a plurality of different scenes, including one or more edge scenes, within a scenario that is at least partially defined by one or more parametric attributes. The computing device also includes a physics engine generate training data representative of the plurality of different scenes including the one or more edge scenes. The physics engine is configured to modify the one or more parametric attributes to generate additional and different training data based upon another plurality of different scenes created by the simulator within another scenario that is at least partially defined by one or more parametric attributes, as modified.

Abstract: Disclosed are various systems and techniques for tracking surgical tools in a surgical video. In one aspect, the system begins by receiving one or more established tracks for one or more previously-detected surgical tools in the surgical video. The system then processes a current frame of the surgical video to detect one or more objects using a first deep-learning model. Next, for each detected object in the one or more detected objects, the system further performs the flowing steps to assign the detected object to a right track: (1) computing a semantic similarity between the detected object and each of the one or more established tracks; (2) computing a spatial similarity between the detected object and the latest predicted location for each of the one or more established tracks; and (3) attempting to assign the detected object to one of the one or more established tracks based on the computed semantic similarity and the spatial similarity metric.

Abstract: An object detection method includes: acquiring an image to be detected; and processing the image to be detected by using an object detection model to obtain an object detection result corresponding to a target object to be detected in the image to be detected. The object detection model includes a feature extraction network and an object detection network. The feature extraction network is used to extract features of the image to be detected to obtain image features related to multiple classes of tar-get objects, and the object detection network is used to process the image features to obtain the object detection result. The object detection network includes a class channel layer, multiple object channel layers and multiple box channel layers.

Abstract: A method includes obtaining a test image captured by a camera of a mobile device, the image including a test component; processing the test image using a machine learning model, wherein the machine learning model has been trained by obtaining a 3D model file representing a 3D model of a training component, generating a plurality of images of the training component based on the 3D model file, the plurality of images differing from one another in at least one of orientation, background, or texture, determining locations of the training component, and training the machine learning model using the plurality of images and the locations of the training component, wherein the machine learning model is trained to detect the training component in photographs captured by mobile devices; and presenting one or more candidate identities of the test component based on an output of the machine learning model.

Abstract: The present invention relates to a method for analysing DNA fragmentation in a sperm cell by approximating the output of a pre-selected chemical assay of sperm DNA fragmentation. According to a first aspect of the present invention, there is provided a method for analysing DNA fragmentation in a sperm cell by approximating the output of a pre-selected chemical assay of sperm DNA fragmentation, the method comprising: providing an image of the sperm cell, under brightfield and/or phase contrast with a total magnification of 400× to 1000×; evaluating the image of the sperm cell to identify and/or measure a pre-selected biomarker; and approximating the output of the pre-selected chemical assay of sperm DNA fragmentation of the sperm cell by subjecting the identified and/or measured biomarker to a first machine learning analysis.

Abstract: In one embodiment, a method includes receiving a querying image associated with a tissue sample after a treatment by a drug compound, identifying a target layer of the tissue sample based on a machine-learning model trained to identify layers of tissue samples, calculating a normalized thickness of the identified target layer, and determining a toxicity indication of the treatment by the drug compound based on the normalized thickness of the identified target layer.

Abstract: Methods and systems for training a model include performing spatial augmentation on an unlabeled input video to generate spatially augmented video. Temporal augmentation is performed on the input video to generate temporally augmented video. Predictions are generated, using a model that was pre-trained on a labeled dataset, for the unlabeled input video, the spatially augmented video, and the temporally augmented video. Parameters of the model are adapted using the predictions while enforcing temporal consistency, temporal consistency, and historical consistency. The model may be used for action recognition in a healthcare context, with recognition results being used for determining whether patients are performing a rehabilitation exercise correctly.

Abstract: Methods, systems, and apparatuses for detecting and describing heterogeneity in a cell sample are disclosed herein. A plurality of fields of view (FOV) are generated for one or more areas of interest (AOI) within an image of the cell sample are generated. Hyperspectral or multispectral data from each FOV is organized into an image stack containing one or more z-layers, with each z-layer containing intensity data for a single marker at each pixel in the FOV. A cluster analysis is applied to the image stacks, wherein the clustering algorithm groups pixels having a similar ratio of detectable marker intensity across layers of the z-axis, thereby generating a plurality of clusters having similar expression patterns.

Abstract: A functional evaluation system of hippocampus that includes: an acquisition unit that acquires image data representing a hippocampus of a subject; and an identification unit that identifies volume data representing a volume of a left hippocampal-amygdaloid transition area of the subject, in accordance with the image data.

Abstract: A method includes obtaining primary image data comprising first image data and second image data; performing labeling on the primary image data, determining the primary image data as training data and determining the labeled image data as validation data, training, based on the training data and the validation data, a first deep learning model where the first image data and the second image data are not distinguished from each other, training, based on a weight of the first deep learning model and based on the training data and the validation data, a second deep learning model where the first image data and the second image data are distinguished from each other according to different views associated with a plurality of cameras, and recognizing, based on the second deep learning model, an object in an image associated with the primary image data, wherein the object is in proximity to the vehicle.

Abstract: An information processing apparatus (1) includes: an acquisition unit (101) configured to acquire an object image that contains one or more objects; a first inference unit (105) configured to infer one or more rectangular areas that respectively surround the one or more objects in the object image, and a type of each of the one or more objects, by applying the object image to a first learning model; a second inference unit (105) configured to infer an area of a target object in the object image by applying the one or more rectangular areas and the type of each of the one or more rectangular areas to a second learning model; and an extraction unit (105) configured to perform color determination on each pixel in the area of the target object and extract colors of the target object.

Abstract: Methods and systems are disclosed for building a few-shot logo recognition system that includes accessing an image with several regions of interest and identifying several objects within the regions of interest using a logo detector neural network. For each object, the logo detector neural network indicates whether the object is a logo. The methods and systems also generate a first and second set of image feature data and a first and second ranked list of logos. A final ranked list of logos is generated based on the first and second ranked list of logos and a category associated with each logo in the final ranked list of logos is identified.

Abstract: Provided are an apparatus for collecting a training image of a deep learning model and a method for the same. The apparatus may include a camera sensor configured to capture an image of a surrounding of a vehicle and a controller configured to determine an entropy of each pixel in the image. The controller may also be configured to determine an entropy of each class in the image based on the entropy of each pixel in the image and determine whether to collect the image based on the entropy of each class.

Abstract: An antiaging assessment method for facial rejuvenation is provided. Firstly, a spatial relationship between various facial key points of an under-test facial image are precisely described. Consequently, a facial feature is obtained. The spatial relationship is an angle relationship or a length relationship. Then, a feature change comparison between the facial feature and reference feature is performed. Consequently, a feature change value is obtained. According to the at least one feature change value, a facial rejuvenation score index is acquired from a corresponding feature change value range of a facial rejuvenation score index data table. The present invention also provides an antiaging assessment device for facial rejuvenation.

Abstract: A computer aided assistance system for use in endoscopic colonoscopy procedures. The computer aided assistance system including: at least one videoendoscopic instrument configured to capture image data; a controller comprising hardware, the controller being connected with the at least one videoendoscopic instrument; and a display connected or integral with the controller, wherein the controller being configured to automatically select a treatment guideline based on a combination of both a size and a classification of a lesion shown in the image data and to display the selected treatment guideline on the display.