Patents.

Systems and methods for safe and reliable autonomous vehicles

Autonomous driving is one of the world's most challenging computational problems. Very large amounts of data from cameras, RADARs, LIDARs, and HD-Maps must be processed to generate commands to control the car safely and comfortably in real-time. This challenging task requires a dedicated supercomputer that is energy-efficient and low-power, with a computational platform that provides diversity and redundancy for functional safety.

Digital credential authentication

A system for creating an identity mapping on a distributed ledger includes an interface and a processor. The interface is configured to receive a request to create an identity mapping on a distributed ledger. The processor is configured to generate an identity key pair; generate a mobile encryption key; encrypt a private identity key of the identity key pair using the mobile encryption key to create an encrypted private key; store the encrypted private key; create a mapping document; sign the mapping document with the private identity key of the identity key pair; and provide the signed mapping document to be stored in a distributed ledger.

Voice attribute conversion using speech to speech

There is provided a computer-implemented method of training a speech-to-speech (S2S) machine learning (ML) model for adapting voice attribute(s) of speech, in which the duration of phones of the second audio content are controlled in response to segment-level durations defined by segment-level start and end time stamps.

Automatic generation of an explanation, based on data lineage and user feedback, of a decision by a machine learning system

A training set for a machine learning system can be modified to improve performance; a message can be transmitted that includes an explanation, derived from data lineage and user feedback, of the system's determination to transmit a particular item of information. (continuation application, not granted)

Determining an effect of a message on a personal brand based on future goals

Provided are techniques for determining an effect of a message on a personal brand based on future goals. A message is received from an entity for a social media platform. Additional messages related to the message are identified. A reaction sentiment trend is determined for each topic of the additional messages for a period of time. An effect the message has on a personal brand of the entity in future is predicted based on future goals of the entity and based on the reaction sentiment trend for each topic. An indication of whether the message is aligned with the future goals based on the predicted effect is provided. In response to the message being aligned with the future goals, the message is posted to the social media platform. In response to the message not being aligned with the future goals, one or more suggestions to modify the message are provided.

Cell-free nucleic acids for the analysis of the human microbiome and components

Systems and methods for optimizing a machine learning model based on a parity metric

Techniques for optimizing a machine learning model. The techniques may include obtaining multiple predictions from a machine learning model, the predictions being based on at least one input feature vector...; creating at least one slice of the predictions based on at least one vector value; determining a sensitive bias metric for the slice based on a sensitive group; determining a base metric for the slice based on a base group; determining a parity metric for the slice based on a ratio of the sensitive bias metric and the base metric; and optimizing the machine learning model based on the parity metric.

Digital credentials for employee badging

A system for credential authentication comprises an interface configured to receive a create indication to create a badge credential representing an employee badge and receive a claim indication from an authentication device to claim the badge credential, and a processor configured to provide the badge credential to the authentication device in response to the claim indication, receive a proof response from the authentication device comprising the badge credential and a lock identifier, validate the proof response using a distributed ledger, and provide a token for unlocking a lock associated with the lock identifier to the authentication device.

Systems and methods for knowledge transfer in machine learning

A training system may create and train a machine learning model with knowledge transfer. The knowledge transfer may transfer knowledge that is acquired by another machine learning model that has been previously trained to the machine learning model that is under training.

Semantic representations using structural ontology for assistant systems

Multitask learning as question answering

Systems and methods for learning unified representations of language, image, and point cloud for three-dimensional recognition

Systems and methods for training a neural network based three-dimensional (3D) encoder for 3D classification are provided. A training dataset including a plurality of samples is received, wherein a first sample includes an image, a text, and a point cloud. An image encoder of a pretrained vision and language model is used to generate image representations for the image of the first sample. A text encoder of the pretrained vision and language model is used to generate text representations for the text of the first sample. The neural network based 3D encoder is used to generate 3D representations for the point cloud of the first sample. A loss objective is computed based on the image representations, text representations, and 3D representations. Parameters of the neural network based 3D encoder are updated based on the computed loss objective via backpropagation.

Systems and methods for learning unified representations of language, image, and point cloud for three-dimensional recognition

A method of training a neural network based three-dimensional (3D) encoder is provided. A training dataset is generated using a plurality of 3D models of a 3D model dataset. To generate a first sample of the training dataset, an image generator with multi-view rendering is used to generate a plurality of image candidates of a first 3D model. A word is chosen from metadata associated with the first 3D model. A language model is used to generate one or more text descriptions using the selected word and a plurality of prompts. A point cloud is generated by randomly sampling points in the 3D model. The first sample is generated to include a first image randomly selected from the plurality of image candidates, one or more text descriptions, and the point cloud is generated. The 3D encoder is trained using the training dataset including the first sample.

Detection of an adversarial backdoor attack on a trained model at inference time

Systems, computer-implemented methods, and computer program products that can facilitate detection of an adversarial backdoor attack on a trained model at inference time are provided. According to an embodiment, a system can comprise a memory that stores computer executable components and a processor that executes the computer executable components stored in the memory. The computer executable components can comprise a log component that records predictions and corresponding activation values generated by a trained model based on inference requests. The computer executable components can further comprise an analysis component that employs a model at an inference time to detect a backdoor trigger request based on the predictions and the corresponding activation values. In some embodiments, the log component records the predictions and the corresponding activation values from one or more layers of the trained model.

Systems and methods for distributed processing of genomic data

Database optimized disaster recovery orchestrator

An example operation may include one or more receiving notifications from one or more monitoring agents, each notification comprising a monitoring agent identifier, one or more configuration changes, and a timestamp corresponding to each configuration change, identifying incremental configuration changes that may require a disaster recovery retest, requesting a partial disaster recovery retest comprising the incremental configuration changes, the partial disaster recovery retest providing test coverage for a subset of a full disaster recovery test plan, and providing a request to a blockchain network to store information for the received notifications to a shared ledger of the blockchain network.

Efficient Scheduling of Build Processes Executing on Parallel Processors

A system schedules tasks using multiple processors. Examples of tasks include build tasks, training of machine learning models, and testing. The system groups the processors into a set of pools. A pool represents a set of processors associated with an affinity class. The system schedules execution of tasks based on the pools.

Image processing neural networks with separable convolutional layers

Efficient private vertical federated learning

Systems and techniques that facilitate universal and efficient privacy-preserving vertical federated learning are provided. In various embodiments, a key distribution component can distribute respective feature-dimension public keys and respective sample-dimension public keys to respective participants in a vertical federated learning framework governed by a coordinator, wherein the respective participants can send to the coordinator respective local model updates encrypted by the respective feature-dimension public keys and respective local datasets encrypted by the respective sample-dimension public keys. In various embodiments, an inference prevention component can verify a participant-related weight vector generated by the coordinator, based on which the key distribution component can distribute to the coordinator a functional feature-dimension secret key that can aggregate the encrypted respective local model updates into a sample-related weight vector. In various embodiments, the inference prevention component can verify the sample-related weight vector, based on which the key distribution component can distribute to the coordinator a functional sample-dimension secret key that can aggregate the encrypted respective local datasets into an update value for a global model.

Semantic representations using structural ontology for assistant systems

Device-agnostic system for planning and executing high-throughput genomic manufacturing operations

Efficient off-policy credit assignment

Pointer Sentinel Mixture Architecture

The technology disclosed provides a so-called “pointer sentinel mixture architecture” for neural network sequence models that has the ability to either reproduce a token from a recent context or produce a token from a predefined vocabulary. In one implementation, a pointer sentinel-LSTM architecture achieves state of the art language modeling performance of 70.9 perplexity on the Penn Treebank dataset, while using far fewer parameters than a standard softmax LSTM.

Systems and methods for optimizing a machine learning-informed automated decisioning workflow in a machine learning task-oriented digital threat mitigation platform

Data privacy protected machine learning systems

Cryptoasset custodial system with different cryptographic keys controlling access to separate groups of private keys

Methods, systems, and apparatus, including medium-encoded computer program products, for secure storage and retrieval of information, such as private keys, useable to control access to a blockchain, include, in at least one aspect, a method including: identifying for an action an associated...

Risk mitigation for a cryptoasset custodial system using data points from multiple mobile devices

Methods, systems, and apparatus for risk mitigation for a cryptoasset custodial system include transmitting an endorsement request for a cryptoasset transaction to a user device configured to cause the user device to prompt a user to endorse the cryptoasset transaction.

Automatic monitoring and adjustment of machine learning model training

Methods and systems for training a machine learning model include training a machine learning model using training data. A status of the machine learning model's training is determined based on an accuracy curve of the machine learning model over the course of the training. Parameters of the training are adjusted based on the status. Training of the machine learning model is completed using the adjusted parameters.

Content based image retrieval for lesion analysis

Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) are commonly used to assess patients with known or suspected pathologies of the lungs and liver. In particular, identification and quantification of possibly malignant regions identified in these high-resolution images is essential for accurate and timely diagnosis. However, careful quantitative assessment of lung and liver lesions is tedious and time consuming. This disclosure describes an automated end-to-end pipeline for accurate lesion detection and segmentation.

Hands-on artificial intelligence education service

Indications of sample machine learning models which create synthetic content items are provided via programmatic interfaces. A representation of a synthetic content item produced by one of the sample models in response to input obtained from a client of a provider network is presented.

Using unstructured input to update heterogeneous data stores

Approaches to using unstructured input to update heterogeneous data stores include receiving unstructured text input, receiving a template for interpreting the unstructured text input, identifying, using an entity classifier, entities in the unstructured text input, identifying one or more potential parent entities from the identified entities based on the template, receiving a selection of a parent entity from the one or more potential parent entities, identifying one or more potential child entities from the identified entities based on the template and the selected parent entity, receiving a selection of a child entity from the one or more potential child entities, identifying an action item in the unstructured text input based on the identified entities and the template, determining, using an intent classifier, an intent of the action item, and updating a data store based on the dete

Method and system for implementing machine learning analysis of documents

Disclosed is an approach for performing auto-classification of documents. A machine learning framework is provided to analyze the document, where labels associated with certain documents can be propagated to other documents.

Access control using a circle of trust

A system for access control includes an interface to receive an access request from a first user application for permission to access a first digital identity wallet application and a processor to: determine whether to grant access for the first user application to the first digital identity wallet application, wherein access is granted for the first user application to the first digital identity wallet application in response to the first user application belonging to a first circle of trust and the first digital identity wallet application belonging to the first circle of trust; and in response to determining to grant access for the first user application to the first digital identity wallet application, provide an access granting indication.

Detecting backdoor attacks using exclusionary reclassification

Embodiments relate to a system, program product, and method for processing an untrusted data set to automatically determine which data points there are poisonous. A neural network is trained network using potentially poisoned training data. Each of the training data points is classified using the network to retain the activations of at least one hidden layer, and segment those activations by the label of corresponding training data. Clustering is applied to the retained activations of each segment, and a clustering assessment is conducted to remove an identified cluster from the data set, form a new training set, and train a second neural model with the new training set. The removed cluster and corresponding data are applied to the trained second neural model to analyze and classify data in the removed cluster as either legitimate or poisonous.

Elastic, ephemeral in-line deduplication service

A deduplication service can be provided to a storage domain from a services framework that expands and contracts to both meet service demand and to conform to resource management of a compute domain. The deduplication service maintains a fingerprint database and reference count data in compute domain resources, but persists these into the storage domain for use in the case of a failure or interruption of the deduplication service in the compute domain. The deduplication service responds to service requests from the storage domain with indications of paths in a user namespace and whether or not a piece of data had a fingerprint match in the fingerprint database. The indication of a match guides the storage domain to either store the piece of data into the storage backend or to reference another piece of data. The deduplication service uses the fingerprints to define paths for corresponding pieces of data.

Server-based game activity suggestions

Digital credentials for step-up authentication

The system comprises an interface and a processor. The interface is configured to receive a request from an application for authorization to access, wherein access to the application is requested by a user, and receive a task request from the application for authorization to access a task, wherein access to the task is requested by the user. The processor is configured to authenticate the request from the application for authorization to access, determine that the task comprises a sensitive task, determine a user authentication device, provide a challenge for a digital credential to the user authentication device, wherein the digital credential is backed by data stored in a distributed ledger, receive a response from the user authentication device, determine the response is valid, and provide an authorization to access the sensitive task.

Identification and application of hyperparameters for machine learning

Methods and systems are provided to determine suitable hyperparameters for a machine learning model and/or feature engineering process. A suitable machine learning model and associated hyperparameters are determined by analyzing a dataset.

Discovering ranked domain relevant terms using knowledge

One embodiment of the invention provides a method for terminology ranking for use in natural language processing. The method comprises receiving a list of terms extracted from a corpus, where the list comprises a ranking of the terms based on frequencies of the terms across the corpus. The method further comprises accessing a domain ontology associated with the corpus, and re-ranking the list based on the domain ontology. The resulting re-ranked list comprises a different ranking of the terms based on relevance of the terms using knowledge from the domain ontology. The method further comprises generating clusters of terms via a trained model adapted to the corpus, and boosting a rank of at least one term of the re-ranked list based on the clusters to increase a relevance of the at least one term using knowledge from the trained model.

Neural network based translation of natural language queries to database queries

Digital credentials for secondary factor authentication

The system for credential authentication comprises an interface and a processor. The interface is configured to receive a request from an application for authorization to access, wherein access to the application is requested by a user using a user device. The processor is configured to provide a login request to the user; validate a login response; determine a user authentication device based on the login response; provide a proof request to the user authentication device; receive a proof response; determine that the proof response is valid using a distributed ledger; generate a token; and provide the token to the application authorizing access for the user.

Global-to-local memory pointer networks for task-oriented dialogue

Structured Cluster Execution for Data Streams

A system for executing a streaming query includes an interface and a processor. The interface is configured to receive a logical query plan. The processor is configured to determine a physical query plan based at least in part on the logical query plan. The physical query plan comprises an ordered set of operators. Each operator of the ordered set of operators comprises an operator input mode and an operator output mode. The processor is further configured to execute the physical query plan using the operator input mode and the operator output mode for each operator of the query.

Server-based video help in a video game

Cryptoasset custodial system using encrypted and distributed client keys

A hardware security module (HSM) generates a client key for an account holder of a cryptoasset custodial system.

Feature engineering in neural networks optimization

A transitive closure data structure is constructed for a pair of features represented in a vector space corresponding to an input dataset. The data structure includes a set of entries corresponding to a set of all possible paths between a first feature in the pair and a second feature in the pair in a graph of the vector space. The data structure is reduced by removing a subset of the set of entries such that only a single entry corresponding to a single path remains in the transitive closure data structure. A feature cross is formed from a cluster of features remaining in a reduced ontology graph resulting from the reducing the transitive closure data structure. A layer is configured in a neural network to represent the feature cross, which causes the neural network to produce a prediction that is within a defined accuracy relative to the dataset.

Cryptoasset custodial system with custom logic

Methods and systems including: receiving a request to take an action in a cryptoasset custodial system for an account holder; authenticating a policy map associated with the action, wherein the policy map defines access control rules governing which actions are allowed under conditions including...

Semantic learning in a federated learning system

A method, a computer system, and a computer program product are provided for federated learning enhanced with semantic learning. An aggregator may receive cluster information from distributed computing devices. The cluster information may relate to identified clusters in sample data of the distributed computing devices. The aggregator may integrate the cluster information to define classes. The integrating may include identifying any redundant clusters amongst the identified clusters. A number of the classes may correspond to a total number of the clusters from the distributed computing devices reduced by any redundant clusters. A deep learning model may be sent from the aggregator to the distributed computing devices. The deep learning model may include an output layer having nodes that may correspond to the defined classes. The aggregator may receive results of federated learning performed by the distributed computing devices. The federated learning may train the deep learning model.

Automatically determining whether an activation cluster contains poisonous data

Embodiments relate to a system, program product, and method for automatically determining which activation data points in a neural model have been poisoned to erroneously indicate association with a particular label or labels. A neural network is trained network using potentially poisoned training data. Each of the training data points is classified using the network to retain the activations of the last hidden layer, and segment those activations by the label of corresponding training data. Clustering is applied to the retained activations of each segment, and a cluster assessment is conducted for each cluster associated with each label to distinguish clusters with potentially poisoned activations from clusters populated with legitimate activations. The assessment includes analyzing, for each cluster, a distance of a median of the activations therein to medians of the activations in the labels.

Systems and methods for unsupervised autoregressive text compression