A method includes generating first contrast significance data for a first computer vision model generated from a first training set of medical scans. First significant contrast parameters are identified based on the first contrast significance data. A first re-contrasted training set is generated based on performing a first intensity transformation function on the first training set of medical scans, where the first intensity transformation function utilizes the first significant contrast parameters. A first re-trained model is generated from the first re-contrasted training set, which is associated with corresponding output labels based on abnormality data for the first training set of medical scans. Re-contrasted image data of a new medical scan is generated based on performing the first intensity transformation function. Inference data indicating at least one abnormality detected in the new medical scan is generated based on utilizing the first re-trained model on the re-contrasted image data.

An apparatus and method to train an image recognition model to accurately estimate a location of a reference point for each class of landmark is disclosed. The apparatus and method use the image recognition model, which is trained based on calculating a class loss and a class-dependent localization loss from training data based on an image recognition model and training the image recognition model using a total loss comprising the class loss and the localization loss.

An attribute recognition system has a person face detection circuitry to detect a suitable person or face for recognition of at least one attribute from persons or faces captured in frame images input from at least one camera to capture a given capture area, an identification information assignment circuitry to identify the persons or faces captured in the frame images having been subjected to the detection by the person face detection circuitry so as to assign an identification information to each identified person or face, and an attribute recognition circuitry to recognize the attribute of a person or face assigned with the identification information, only if the person or face is yet without being subjected to recognition of the attribute, and at the same time if the person or face has been detected by the person face detection circuitry as a suitable person or face for the recognition of the attribute.

An attribute recognition system has a person face detection circuitry to detect a suitable person or face for recognition of at least one attribute from persons or faces captured in frame images input from at least one camera to capture a given capture area, an identification information assignment circuitry to identify the persons or faces captured in the frame images having been subjected to the detection by the person face detection circuitry so as to assign an identification information to each identified person or face, and an attribute recognition circuitry to recognize the attribute of a person or face assigned with the identification information, only if the person or face is yet without being subjected to recognition of the attribute, and at the same time if the person or face has been detected by the person face detection circuitry as a suitable person or face for the recognition of the attribute.

Disclosed herein are a visual story generation method and apparatus for generating a story from a plurality of images by using a deep learning network. The visual story generation method includes: extracting features from a plurality of respective images by using the first extraction unit of a deep learning network; generating the structure of a story based on the overall feature of the plurality of images by using the second extraction unit of the deep learning network; and generating the story by using outputs of the first and second extraction units.

Disclosed herein are a visual story generation method and apparatus for generating a story from a plurality of images by using a deep learning network. The visual story generation method includes: extracting features from a plurality of respective images by using the first extraction unit of a deep learning network; generating the structure of a story based on the overall feature of the plurality of images by using the second extraction unit of the deep learning network; and generating the story by using outputs of the first and second extraction units.

A computer architecture may comprise a processor, a memory, and a differential memory subsystem (DMS). A learning engine is stored on the memory and configured to present data to an expert user, to receive user sensory input measuring reactions related to the presented data, and to create an attention map based thereon. The attention map is indicative of portions of the presented data on which the expert user focused. The learning engine is configured to annotate the attention map with the natural language input labels and to train a neural network based on the user sensory input. The learning engine is configured to create a model based on the trained neural network, to provide an application program for an output target; and to instruct the output target via the application program to detect and remedy anomalous activity. The DMS is physically separate and configured for experimental data processing functions.

In various examples, sensor data may be received that represents a field of view of a sensor of a vehicle located in a physical environment. The sensor data may be applied to a machine learning model that computes both a set of boundary points that correspond to a boundary dividing drivable free-space from non-drivable space in the physical environment and class labels for boundary points of the set of boundary points that correspond to the boundary. Locations within the physical environment may be determined from the set of boundary points represented by the sensor data, and the vehicle may be controlled through the physical environment within the drivable free-space using the locations and the class labels.

In various examples, sensor data may be received that represents a field of view of a sensor of a vehicle located in a physical environment. The sensor data may be applied to a machine learning model that computes both a set of boundary points that correspond to a boundary dividing drivable free-space from non-drivable space in the physical environment and class labels for boundary points of the set of boundary points that correspond to the boundary. Locations within the physical environment may be determined from the set of boundary points represented by the sensor data, and the vehicle may be controlled through the physical environment within the drivable free-space using the locations and the class labels.

A convolutional neural network (CNN) and associated method for identifying basal cell carcinoma are disclosed. The CNN comprises two convolution layers, two pooling layers and at least one fully-connected layer. The first convolution layer uses initial Gabor filters that model the kernel parameters setting in advance based on human professional knowledge. The method uses collagen fiber images for training images and converts doctors' knowledge to initiate the Gabor filters as featuring computerization. The invention provides better training performance in terms of training time consumption and training material overhead.