A computing device, method and computer program product are provided to generate synthetic images of abnormalities on the surface of an object, such as a vehicle. The synthetic images of abnormalities on the surface of an object may be utilized for training a machine learning algorithm to detect and/or classify abnormalities. In the context of a method, a respective abnormality is parametrically modeled by selecting one or more control points that satisfy parameters associated with the respective abnormality and generating a surface representative of the respective abnormality based on the one or more control points. The method also renders a synthetic image of at least a portion of the surface of the object having the respective abnormality as defined by the parametric modeling thereof. The rendering of the synthetic image includes rendering the synthetic image in accordance with a predefined lighting condition and from a predefined viewpoint.

A computer system (which may include one or more computers) that trains a neural network is described. During operation, the computer system may train the neural network based at least in part on a set of hyperparameters, where the training includes computing weights associated with neurons in the neural network. Moreover, during the training, the computer system may dynamically adapt one or more first hyperparameters in the set of hyperparameters based at least in part on a measure corresponding to a local geometry of a loss landscape at or proximate to a current location in the loss landscape. Note that the dynamic adapting based at least in part on the measure is separate from or in addition to a predefined adaptation of one or more second hyperparameters the set of hyperparameters based on a predefined number of iterations or cycles in the training or a predefined scaling factor.

Methods, systems, and computer-readable storage media for receiving input data including a set of entities of a first type and a set of entities of a second type, providing a set of features based on entities of the first type, the set of features including features expected to be included in entities of the second type, filtering entities of the second type based on the set of features to provide a sub-set of entities of the second type, and generating an output by processing the set of entities of the first type and the sub-set of entities of the second type through a ML model, the output comprising a set of matching pairs, each matching pair in the set of matching pairs comprising an entity of the set of entities of the first type and at least one entity of the sub-set of entities of the second type.

One embodiment provides a method, including: receiving a dataset and a model corresponding to a bias checker, wherein the bias checker detects bias within both the dataset and the model, based upon a bias checking algorithm and a bias checking policy, wherein the dataset comprises a plurality of attributes; testing the bias checking algorithm of the bias checker by (i) generating test cases that modify the dataset by introducing bias therein and (ii) running the bias checker against the modified dataset; testing the bias checking policy of the bias checker by generating a plurality of test cases and running the bias checker against the plurality of test cases; and providing a notification to a user regarding whether the bias checker failed to indicate bias for one or more of the plurality of attributes.

Automated vehicle repair estimation by aggregate ensembling of multiple artificial intelligence functions is provided. A method comprises receiving a plurality of vehicle repair recommendation sets for a damaged vehicle, wherein each of the vehicle repair recommendation sets identifies at least one recommended vehicle repair operation of a plurality of the vehicle repair operations for the damaged vehicle; aggregating a plurality of the recommended vehicle repair operations; generating a composite vehicle repair recommendation set that identifies the aggregated recommended vehicle repair operations; and providing the composite vehicle repair recommendation set to one or more vehicle repair insurance claims management systems.

Aspects include receiving a request to perform an image classification task in a target domain. The image classification task includes identifying a feature in images in the target domain. Classification information related to the feature is transferred from a source domain to the target domain. The transferring includes receiving a plurality of pairs of task-irrelevant images that each includes a task-irrelevant image in the source domain and in the target domain. The task-irrelevant image in the source domain has a fixed correspondence to the task-irrelevant image in the target domain. A target neural network is trained to perform the image classification task in the target domain. The training is based on the plurality of pairs of task-irrelevant images. The image classification task is performed in the target domain and includes applying the target neural network to an image in the target domain and outputting an identified feature.

Provided is a learning-based data processing system which generates a learning model by learning a learning data set, recognizes observational data according to the learning model, and provides a recognition result. The learning-based data processing system may include a data recognition device configured to generate a cascaded learning model by cascading a first learning model generated based on a first learning data set and a second learning model generated based on a second learning data set.

A method for sorting geographic location points, a method for training a sorting model and corresponding apparatuses are disclosed, which relates to the technical field of big data. A specific implementation solution is: receiving a query request for geographic location points of a vertical class from a user; inputting candidate geographic location point data of the vertical class into a preference model of the user, to obtain a preference score of the user for each candidate geographic location point; inputting the preference score of the user for each candidate geographic location point into a sorting model as one of sorting features of each candidate geographic location point, to obtain a sorting score of each candidate geographic location point; and determining, according to the sorting score of each candidate geographic location point, a query result returned to the user. The present disclosure can integrate preference factors of a user into sorting when the user queries geographic location points of a vertical class, so that query results can meet the user's personalized needs.

The present disclosure describes transaction-enabling systems and methods. A system can include a facility including a core task including a customer relevant output and a controller. The controller may include a facility description circuit to interpret a plurality of historical facility parameter values and corresponding facility outcome values and a facility prediction circuit to operate an adaptive learning system, wherein the adaptive learning system is configured to train a facility production predictor in response to the historical facility parameter values and the corresponding outcome values. The facility description circuit also interprets a plurality of present state facility parameter values, wherein the trained facility production predictor determines a customer contact indicator in response to the plurality of present state facility parameter values and a customer notification circuit provides a notification to a customer in response.

A vibration-based authentication method for an access control system includes: collecting vibration signals generated by a built-in vibration motor in an authentication device; filtering, denoising, and performing endpoint segmentation on the collected vibration signals, and extracting vibration signals containing effective touch; performing an alignment on the segmented vibration signals; performing a fast Fourier transform on the aligned vibration signals to obtain frequency-domain data, extracting frequency-domain features obtained after alignment and features obtained before alignment to construct a training data set, and storing the training data set in a database of the authentication device; using a new unlock signal generated when a user touches the authentication device as test data, and processing the test data to obtain test data containing effective touch; and matching and classifying the test data containing effective touch with the training data set by using a machine learning classification model, to obtain an authentication result.