Abstract: Systems and methods for censoring text characters in text-based data are provided. In some embodiments, an artificial intelligence system may be configured to receive text-based data and store the text-based data in a database The artificial intelligence system may be configured to receive a list of target pattern types identifying sensitive data and receive censorship rules for the target pattern types determining target pattern types requiring censorship.

Abstract: Systems and methods for determining neural network brittleness are disclosed. For example, the system may include one or more memory units storing instructions and one or more processors configured to execute the instructions to perform operations. The operations may include receiving a modeling request comprising a preliminary model and a dataset. The operations may include determining a preliminary brittleness score of the preliminary model. The operations may include identifying a reference model and determining a reference brittleness score of the reference model. The operations may include comparing the preliminary brittleness score to the reference brittleness score and generating a preferred model based on the comparison. The operations may include providing the preferred model.

Abstract: Methods and systems described herein relate to authenticating users-based on generating images that have modified features. More specifically, the methods and systems generate these images by processing existing training data to identify a common feature in existing photos that may be modified, and then modifying that feature with the use of generative adversarial networks.

Abstract: A method for generating a cluster-based machine learning model based on federated learning with cluster feedback includes providing a current machine learning model to a plurality of user devices that train the current machine learning model, receiving respective model states, generating updated model states, causing the plurality of user devices to obtain a respective instance of an updated machine learning model based on the updated model states, receiving an applicability feedback for the updated machine learning model for each of the plurality of user devices, determining a plurality of user clusters including a subset of the plurality of user devices, identifying a first user cluster and a second user cluster, the first user cluster having a higher cluster applicability feedback than the second user cluster, receiving the additional model states from the clusters and updating the updated machine learning model to generate the cluster-based machine learning model.

Abstract: Systems and methods for generating synthetic intercorrelated data are disclosed. For example, a system may include at least one memory storing instructions and one or more processors configured to execute the instructions to perform operations. The operations may include training a parent model by iteratively performing steps. The steps may include generating, using the parent model, first latent-space data and second latent-space data. The steps may include generating, using a first child model, first synthetic data based on the first latent-space data, and generating, using a second child model, second synthetic data based on the second latent-space data. The steps may include comparing the first synthetic data and second synthetic data to training data.

Abstract: The present disclosure provides systems and methods for synthetic data generation. A recurrent neural network can be trained for synthetic data generation by obtaining a sequence of elements and determining, using a classifier, that the sequence corresponds to a token. In response to the determination, a recurrent neural network configured to use a first vocabulary including the elements can be modified to use a second vocabulary, the second vocabulary including the token and the first vocabulary. The modified recurrent neural network can be trained using the token and the sequence of elements. The trained recurrent neural network can be used to generate synthetic data. A classifier can detect sequences of elements in the synthetic data corresponding to tokens. The tokens can replace the sequences of elements in the generated synthetic data and can be provided to the trained recurrent neural network to continue synthetic data generation.

Abstract: Systems and methods for generating real-time location-based notifications are described. In some aspects, the system processes, using a first machine learning model, a real-time video stream to determine a measure of significance of an event depicted in the real-time video stream. The system determines whether the user was present at the event. Based on determining that the user was present at the event, the system processes a user profile for the user and the measure of significance of the event to determine whether to generate for the user a cryptographic token associated with the event. Based on determining to generate for the user the cryptographic token associated with the event, the system generates a notification to the user including information for accessing the cryptographic token for the user.

Abstract: Aspects described herein may allow for deploying third-party callback to be performed on a mobile device. For example, the mobile device may configure a third-party callback to be performed after a triggering event is detected by a remote device. The triggering event may relate to a remote device associated with the mobile device. A determination may be made as to whether the mobile device is within a pre-determined proximity of the remote device. If the two devices are in proximity and if the mobile device receives a message that indicates the triggering event has been detected, the mobile device may perform the third-party callback. In this way, a user's need to request services conveniently and accurately is facilitated.

Abstract: A system for determining data requirements to generate machine-learning models. The system may include one or more processors and one or more storage devices storing instructions. When executed, the instructions may configure the one or more processors to perform operations including: receiving a sample dataset, generating a plurality of data categories based on the sample dataset; generating a plurality of primary models of different model types using data from the corresponding one of the data categories as training data; generating a sequence of secondary models by training the corresponding one of the primary models with progressively less training data; identifying minimum viable models in the sequences of secondary models; determining a number of samples required for the minimum viable models; and generating entries in the database associating: model types; corresponding data categories; and corresponding numbers of samples in the training data used for the minimum viable models.

Abstract: Methods and systems disclosed herein describe using machine learning to lock and unlock a device. Machine learning may be trained to recognize one or more features. Once the device has been trained to recognize one or more features, a user may define an unlock condition for the device using the one or more trained features. After defining the unlock condition, the device may be locked by verifying the one or more features that the user defined as the unlock condition using machine learning. When verification is successful, the device may be unlocked and the user allowed to access the device.

Abstract: Systems and methods for architecture embeddings for efficient dynamic synthetic data generation are disclosed. The disclosed systems and methods may include a system for generating synthetic data configured to perform operations. The operations may include retrieving a set of rules associated with a first data profile and generating, by executing a hyperparameter search, a plurality of hyperparameter sets for generative adversarial networks (GANs) that satisfy the set of rules. The operations may include generating mappings between the hyperparameter sets and the first data profile and storing the mappings in a hyperparameter library. The operations may include receiving a request for synthetic data, the request indicating a second data profile and selecting, from the mappings in the hyperparameter library, a hyperparameter set mapped to the second data profile. The operations may include building a GAN using the selected hyperparameter set and generating, using the GAN, a synthetic data set.

Abstract: Aspects described herein may relate to methods, systems, and apparatuses for labeling data in connection with synthetic data generation. The data labeling may begin with a manual process where a user provides labels for data. Based on the labels provided by the user, modified data may be generated and may include one or more encodings associated with the labels provided by the user. A machine-learning model may be trained to predict labels based on the modified data samples. Accuracy of the model may be determined based on comparing the predicted labels to further labels provided by the user and/or by allowing the user to indicate whether predicted labels are correct or incorrect. Once the model is determined to be accurate, the predicted labels may be used as a basis for generating synthetic data.

Abstract: Disclosed embodiments may include a method for authentication using partitioned authentication tokens. The system can receive an indication of a first and second user device associated with a user. The indication can include a priority order of the first and second user device. The system can then receive an authentication request associated with the user from an application. The system can generate an authentication token to authenticate the user and partition the authentication token to create a first token portion and a second token portion. The system can determine which device of the first and second user device has a higher priority based on the priority order and can transmit the first token portion and the second token portion to the devices in order of priority. The system can receive a receipt of the token portions and transmit instructions to the application to authenticate the user.

Abstract: A method for training a neural network model includes generating a training dataset with a plurality of data types and word samples belonging to each data type. A plurality of character strings stored in a plurality of data fields in a first data file are received where the plurality of character strings includes at least one word belonging to at least one data type in the plurality of data types. The at least one word from each of the plurality of character strings in each of the data fields are split and matched to the at least one data type using the neural network model. An ad hoc second data file with a plurality of data vectors is constructed based on a user selection of data field labels where each data vector includes words matched to a data type with a respective data field label.

Abstract: Systems as described herein may label data to preserve privacy. An annotation server may receive a document comprising a collection of text representing a plurality of confidential data from a first computing device. The annotation server may convert the document to a plurality of text embeddings. The annotation server may input the text embeddings into a machine learning model to generate a plurality of synthetic images, and receive a label for each of the plurality of synthetic images from a third-party labeler. Accordingly, the annotation server may send the confidential data and the corresponding labels to a second computing device.

Abstract: Methods and systems for the dynamically re-clustering of nodes in clusters to provide optimal performance and/or the most efficient use of resources through the use of machine learning models. Specifically, the methods and systems may determine a cluster that optimally performs and/or has the most efficient use of resources based on a first machine learning model. The methods and system may then retrieve available substitute nodes from other domains and/or networks that may lie outside the cluster, but may nonetheless be available to, or accessed by the cluster. The methods and systems may then generate an additional plurality of clusters using one or more of the original nodes of the originally selected clusters and/or one or more of the available substitute nodes.

Abstract: Systems and methods described herein discuss securing user-entered data in-transit between a first device and a second device. A user may enter text in a document. A first device may analyze the document to identify the user-entered text. The user-entered text may be separated from the document and transformed into an image using a machine learning algorithm. Transforming the text into an image may secure the data in-transit from the first device to a second device. The second device may receive the image and the document from the first device. The second device may reconstruct the user-entered text from the received image and re-assemble the document from the received document and the reconstructed user-entered text.

Abstract: In some embodiments, gaze-tracking-based image downscaling for multi-party video communication may be provided. In some embodiments, a set of gaze locations may be received from a set of receiving devices during concurrent video communication sessions between the receiving devices and a sending device. Different collections of positions may be determined based on the gaze locations, and, for each such collection, a spatial indicator (e.g., a bounded region) may be determined. A first downscaled encoding of the source image (from the sending device) may be generated based on the first spatial indicator, and a second downscaled encoding of the source image may be generated based on the second spatial indicator, etc. The downscaled encodings may then be sent to the respective receiving devices during the concurrent video communication sessions.

Abstract: Disclosed embodiments may include systems and methods for utilizing gesture-based transaction cards. Examples include: first, using a mobile device to preauthorize transaction cards to enter an unlocked state before making a purchase by making a gesture with the transaction card toward the mobile device. Second, using a first transaction card to change an authorization condition of a second transaction card associated with the same account by making a gesture from the first transaction card to the second transaction card. Third, using a first transaction card to distribute a predebited virtual card to the user device of a second user by making a gesture from the first transaction card to the user device of the second user. Additionally, the primary user may be able to upload new authorization conditions to the transaction card and control the system through a related mobile application on a connected mobile device.

Abstract: Disclosed embodiments may include systems and methods for utilizing gesture-based transaction cards. Examples include: first, using a mobile device to preauthorize transaction cards to enter an unlocked state before making a purchase by making a gesture with the transaction card toward the mobile device. Second, using a first transaction card to change an authorization condition of a second transaction card associated with the same account by making a gesture from the first transaction card to the second transaction card. Third, using a first transaction card to distribute a predebited virtual card to the user device of a second user by making a gesture from the first transaction card to the user device of the second user. Additionally, the primary user may be able to upload new authorization conditions to the transaction card and control the system through a related mobile application on a connected mobile device.