A computer-implemented method for implementing face recognition includes receiving training data including a plurality of augmented images each corresponding to a respective one of a plurality of input images augmented by one of a plurality of variations, splitting a feature embedding generated from the training data into a plurality of sub-embeddings each associated with one of the plurality of variations, associating each of the plurality of sub-embeddings with respective ones of a plurality of confidence values, and applying a plurality of losses including a confidence-aware identification loss and a variation-decorrelation loss to the plurality of sub-embeddings and the plurality of confidence values to improve face recognition performance by learning the plurality of sub-embeddings.

Methods, apparatus, systems and articles of manufacture are disclosed for classification of unknown samples using agglomerative clustering. An apparatus includes an extractor to extract a feature from a sample source code, the feature including at least one of a register, a variable, or a library based on a threshold of occurrence in a corpus of samples, the corpus of samples including malware samples, a dendrogram generator to generate a dendrogram based on features extracted from the sample source code, the dendrogram representing a collection of samples clustered based on similarity among the samples, the samples including sample clusters belonging to known malware families, and an anchor point identifier to traverse the dendrogram to identify similarity of an unknown sample to the sample clusters based on a confidence score, and identify anchor point samples from the sample clusters identified as similar to the unknown sample, the anchor point samples to provide metadata for use in extrapolating information to classify the unknown sample.

According to a first aspect, it is presented a method for dynamically selecting a learning model for a sensor device. The learning model is configured for determining output data based on sensor. The method comprises the steps of: detecting a need for a new learning model for the sensor device based on performance of a currently loaded learning model in the sensor device; determining at least one feature candidate based on sensor data from the at least one sensor, wherein each one of the at least one feature candidate is associated with a different source of sensor data; selecting a new learning model, from a set of candidate learning models, based on the at least one feature candidate and input features of each one of the candidate learning models; and triggering the new learning model to be loaded on the sensor device, replacing the currently loaded learning model.

Systems, methods, and computer-readable media for achieving privacy for both data and an algorithm that operates on the data. A system can involve receiving an algorithm from an algorithm provider and receiving data from a data provider, dividing the algorithm into a first algorithm subset and a second algorithm subset and dividing the data into a first data subset and a second data subset, sending the first algorithm subset and the first data subset to the algorithm provider and sending the second algorithm subset and the second data subset to the data provider, receiving a first partial result from the algorithm provider based on the first algorithm subset and first data subset and receiving a second partial result from the data provider based on the second algorithm subset and the second data subset, and determining a combined result based on the first partial result and the second partial result.

Certain aspects of the present disclosure provide techniques for performing piecewise pointwise convolution, comprising: performing a first piecewise pointwise convolution on a first subset of data received via a first branch input at a piecewise pointwise convolution layer of a convolutional neural network (CNN) model; performing a second piecewise pointwise convolution on a second subset of data received via a second branch input at the piecewise pointwise convolution layer; determining a piecewise pointwise convolution output by summing a result of the first piecewise pointwise convolution and a result of the second piecewise pointwise convolution; and providing the piecewise pointwise convolution output to a second layer of the CNN model.

Provided are an object identification method, apparatus, and device. The object identification method comprises: acquiring a first image of at least part of an object; determining a feature portion of the object on the basis of the first image; acquiring a second image of the feature portion of the object; and identifying an object category of the object on the basis of the second image. According to the object identification method, apparatus, and device, in which a feature portion of an object is acquired and identification of the category of the object is performed on the basis of the feature portion, operations are simple, and the accuracy of object identification can be effectively improved.

In some implementations, a device may obtain a spectroscopic measurement associated with a sample. The device may generate, based on the spectroscopic measurement and a global classification model, a local classification model that includes a plurality of classes. The device may identify, based on the spectroscopic measurement, a particular class of the plurality of classes of the local classification model. The device may identify a prediction threshold associated with the particular class. The device may classify, based on the particular class and the prediction threshold, the spectroscopic measurement. The device may provide, based on classifying the spectroscopic measurement, information indicating whether the sample belongs to the particular class.

A system for generating a contention scheme includes a computing device, the computing device configured to obtain a solvency signature as a function of a solvency entity, determine a solvency grouping as a function of the solvency signature, identify a null element as a function of the solvency grouping, wherein identifying the null element further comprises receiving a regulation element as a function of a regulation database, and identifying the null element as a function of the regulation element and the solvency grouping, produce a weighted vector as a function of the null element, and generate a contention scheme as a function of the weighted vector.

A method, system, and computer program product for visualizing a machine learning model are provided. A confusion matrix and model performance metric data are received from a classification model. For each data point in the confusion matrix, a corresponding pixel is generated. The pixels are grouped into clusters. Each cluster represents a label in the confusion matrix. A centroid is generated for each cluster. Using the model performance metric data, a misclassification indicator arrow is generated for each misclassified data point. The misclassification indicator arrow indicates both the predicted class and the actual class. The clusters, the centroids, and the misclassification indicator arrow are displayed as a graphical visualization of the machine learning model.

A system includes: a light source that emits pulsed light that illuminates a user's head portion; a photodetector that detects at least part of pulsed light returning from the head portion and that outputs one or more signals corresponding to an intensity of the at least part; electrical circuitry; and a memory that stores an emotion model indicating a relationship between the one or more signals and emotions. Based on a change in the one or more signals, the electrical circuitry selects an emotion by referring to the model. The one or more signals include a first signal corresponding to an intensity of first part of the reflection pulsed light and a second signal corresponding to an intensity of second part of the reflection pulsed light. The first part includes part before a falling period is started; and the second part includes at least part in the falling period.