A server for classifying an image and a method of operating the server are provided. The method includes obtaining classification probability values of the image, by applying the image to an image classification model including a plurality of parallel multi-layer perceptron (MLP) layers, and classifying the image, based on the classification probability values, wherein each of the plurality of parallel MLP layers includes a first MLP and a second MLP, an operation using the first MLP and an operation using the second MLP are performed in parallel, and dimensions of data before and after an operation of each of the plurality of parallel MLP layers are same by combining an operation result of the first MLP with an operation result of the second MLP.

In an object recognition device, an image acquisition means acquires an image. A recognition means recognizes each of objects included in an image, and generates each recognition result. A graph generation means defines each of the recognized objects on one of a node and an edge and defines each of relationships among the objects on another one of the node and the edge, based on each recognition result. A graph analysis means analysis a graph and generates an analysis result indicating the relationships among the objects.

In a point cloud data set represented by an image, a first object can be caused to be distinguished from a second object. Current positions of points in the point cloud data set can be arguments for functions of a set of functions. The set of functions can include attraction functions for the points and repulsion functions for the points. Results of the set of functions can be calculated. Based on the results, at least some of the points can be caused to move from the current positions to new positions. In the point cloud data set represented by the image with the at least some of the points being at the new positions, the first object can be distinguished from the second object.

Disclosed is method for order picking, method comprising: obtaining image frame captured by camera arranged on a pallet jack; processing image frame, using object detection model that is pre-trained, for generating first list, wherein said model identifies and localizes case(s) represented in image frame, first list includes one entry per image frame; processing image segment(s) representing case(s), using classification model that is pre-trained, for updating first list, classification model classifies product in given case as belonging to given class, predicts confidence score and generates identification code of product, first list is updated by adding confidence score and identification code to given entry; employing tracking algorithm for generating tracking list indicating count of cases picked per product for pallet, tracking algorithm utilizes first list; and providing, on display device, interactive user interface, in real time, for presenting count of cases picked per product for pallet.

The present disclosure provides feature data processing methods and devices. One exemplary feature data processing method comprises: classifying features into an important feature set and an auxiliary feature set according to information attribute values of the features; converting features in the auxiliary feature set to hash features; and combining the hash features with features in the important feature set, and setting the combined features as fingerprint features. Training and prediction of to-be-processed data can be performed based on the fingerprint features. With the embodiments of the present disclosure, training dimensions can be more controllable and the amount training data amount can be reduced. Therefore, the efficiency of data processing can be improved.

An embodiment of an image processor device includes technology to fetch a feature point data set from outside a local memory, locally store three or more fetched feature point data sets in the local memory, compute orientation information for each fetched feature point data set, compute first descriptor information based on the computed orientation information and a first locally stored feature point data set in parallel with a fetch and local store of a second feature point data set in the local memory, and compute second descriptor information based on the computed orientation information and the second locally stored feature point data set in parallel with the compute of the first descriptor information. Other embodiments are disclosed and claimed.

A system for prioritizing user symptom complaint inputs includes a computing device, wherein the computing device is configured to receive a plurality of symptom complaint datums generated by a user, determine a frequency element as a function of the plurality of symptom complaint datums, produce a disease criticality score as a function of the plurality of the frequency element, wherein producing further comprises obtaining an expert input, and determining the criticality score as a function of the expert input and the frequency element, and generate a suspected disease state as a function of the disease criticality score.

The present invention relates to a method, device, and robot for classifying an outlier during object recognition learning using artificial intelligence. The method or device for classifying an outlier during object recognition learning according to an embodiment of the present invention sets an inlier region and an outlier region through learning using unlabeled data and labeled data.

Methods of segmenting an abdominal image, computer apparatuses and storage mediums. The method includes acquiring an abdominal image to be examined; and classifying pixels in the abdominal image to be examined based on a trained full convolution neural network, and determining a segmented image corresponding to the abdominal image to be examined, wherein the trained full convolution neural network is trained and determined based on a first training set and a second training set, the first training set includes first sample abdominal images and pixel classification label images corresponding to the first sample abdominal images, the second training set includes second sample abdominal images and the number of pixels of second sample abdominal images correspondingly belong to each class. Through the methods herein, the accuracy of the segmentation can be improved.

A computer-implemented method of training an object detector, the method comprising: training an embedding neural network using, as an input, cropped images from an image dataset, wherein training the embedding neural network is performed using a self-supervised learning approach and the trained embedding neural network translates input images into a lower dimensional representation; and training an object detector neural network by, for images of the image dataset, repeatedly: passing an image through the object detector neural network to obtain proposed coordinates of an object within the image, cropping the image to the proposed coordinates to obtain a cropped image, passing the cropped image through the trained embedding neural network to obtain a cropped image representation, passing an exemplar through the trained embedding neural network to obtain an exemplar representation, wherein the exemplar is a cropped manually labelled image bounding a known object, computing a distance in embedding space between the cropped image representation and the exemplar representation, computing a gradient of the cropped image representation and the exemplar representation with respect to the distance, and passing the gradient into the object detector neural network for use in backpropagation to optimise the object detector neural network.