Techniques are disclosed for training machine learning systems. An input device receives training data comprising pairs of training inputs and training labels. A generative memory assigns training inputs to each archetype task of a plurality of archetype tasks, each archetype task representative of a cluster of related tasks within a task space and assigns a skill to each archetype task. The generative memory generates, from each archetype task, auxiliary data comprising pairs of auxiliary inputs and auxiliary labels. A machine learning system trains a machine learning model to apply a skill assigned to an archetype task to training and auxiliary inputs assigned to the archetype task to obtain output labels corresponding to the training and auxiliary labels associated with the training and auxiliary inputs assigned to the archetype task to enable scalable learning to obtain labels for new tasks for which the machine learning model has not previously been trained.

The present disclosure provides systems and methods for providing augmented reality experiences. Consistent with disclosed embodiments, one or more machine-learning models can be trained to selectively process image data. A pre-processor can be configured to receive image data provided by a user device and trained to automatically determine whether to select and apply a preprocessing technique to the image data. A classifier can be trained to identify whether the image data received from the pre-processor includes a match to one of a plurality of triggers. A selection engine can be trained to select, based on a matched trigger and in response to the identification of the match, a processing engine. The processing engine can be configured to generate an output using the image data, and store the output or provide the output to the user device or a client system.

A method of data augmentation is provided. The method includes the following operations: selecting an original image from an original dataset including label data configured to indicate a labeled area of the original image; selecting at least part of content, located in the labeled area, of the original image as a first target image; generating a first sample image according to the first target image, in which the first sample image includes the first target image and a first border pattern different from the first target image, and the content, located in the labeled area, of the original image is free from including at least part of the first border pattern; and incorporating the first sample image into a sample dataset, in which the sample dataset is configured to be inputted to a machine learning model.

In a creation method of a trained model, a reconstructed image (60) obtained by reconstructing three-dimensional X-ray image data (80) is generated. A projection image (61) is generated from a three-dimensional model of an image element (50) by a simulation. The projection image is superimposed on the reconstructed image to generate a superimposed image (67). A trained model (40) is created by performing machine learning using the superimposed image, and the reconstructed image or the projection image.

Systems and methods are provided for analyzing messages generated by a plurality of computing devices associated with a plurality of users in a messaging system to generate training data to train a machine learning model to determine a probability that a media content item was generated inside an enclosed location or outside, receiving a media content item from a computing device, analyzing the media content item using the trained machine learning model to determine a probability that the media content item was generated inside an enclosed location or outside, determining, based on the probability generated by the trained machine learning model, that the media content item was generated inside an enclosed location, and determining an inside temperature associated with the venue based on messages generated by a plurality of computing devices in a messaging system comprising media content items and temperature information for the venue or a similar venue type.

Embodiments described herein provide for training a machine learning model for automatic organ segmentation. A processor executes a machine learning model using an image to output at least one predicted organ label for a plurality of pixels of the image. Upon transmitting the at least one predicted organ label to a correction computing device, the processor receives one or more image fragments identifying corrections to the at least one predicted organ label. Upon transmitting the one or more image fragments and the image to a plurality of reviewer computing devices, the processor receives a plurality of inputs indicating whether the one or more image fragments are correct. When a number of inputs indicating an image fragment of the image fragments is correct exceeds a threshold, the processor aggregates the image fragment into a training data set. The processor trains the machine learning model with the training data set.

A method for generating a graph structure for training a graph neural network. The method includes: obtaining data representing a computational graph, wherein the computational graph comprises a plurality of nodes connected by edges; and generating the graph structure for training the graph neural network by removing edges from the computational graph. The edges are removed in such a way that an environment in the computational graph corresponds to an environment in the graph structure.

A computer-program product storing instructions which, when executed by a computer, cause the computer to, for one or more iterations, update parameters associated with a machine-learning network utilizing perturbations for input data, wherein the perturbations are sampled utilizing Markov chain Monte Carlo, identify a loss value associated with each perturbation in each iteration, and evaluate the machine learning network by identifying an average loss value across each iteration and outputting the average loss value.

A system and method for adaptive color assignments to image labels during annotation of datasets includes preparing a dataset for image labeling by an annotator by: leveraging a global color analyzer to perform a global color distribution of a plurality of images to identify one or more overall colors present in the plurality of images, and a local color analyzer to perform a local color distribution for each image to identify one or more colors present in an area of interest of the image, and selecting a plurality of candidate colors to be used as image labels by the annotator, based on an output of the global color analyzer and an output of the local color analyzer.

Disclosed is an energy-efficient retraining method of a generative neural network for domain-specific optimization, including (a) retraining, by a mobile device, a pretrained generative neural network model with respect to some data of a new user dataset, (b) comparing, by the mobile device, the pretrained generative neural network model and a generative neural network model retrained for each layer with each other in terms of a relative change rate of weights, (c) selecting, by the mobile device, specific layers having high relative change rate of weights, among layers of the pretrained generative neural network model, as layers to be retrained, and (d) performing, by the mobile device, weight update for only the layers selected in step (c), wherein only some of all layers are selected and trained in a retraining process that requires a large amount of operation, whereby rapid retraining is performed in the mobile device.