Neural networks

ASSESSMENT OF MACHINE-LEARNING ALGORITHM

Absstract of: EP4528555A1

Techniques are disclosed that enable assessment of the processing of datasets in a machine-learning algorithm such as a deep convolutional neural network. Layer-based assessment is possible. The processing of images can be assessed. Explainable artificial intelligence is possible. Safe control of autonomous vehicles is possible.

TECHNIQUES FOR ARTIFICIAL INTELLIGENCE CAPABILITIES AT A NETWORK SWITCH

Absstract of: US2025097120A1

Examples include techniques for artificial intelligence (AI) capabilities at a network switch. These examples include receiving a request to register a neural network for loading to an inference resource located at the network switch and loading the neural network based on information included in the request to support an AI service to be provided by users requesting the AI service.

System and method for orientating capture of ultrasound images

Absstract of: AU2025201525A1

SYSTEM AND METHOD FOR ORIENTATING CAPTURE OF ULTRASOUND IMAGES A downloadable navigator for a mobile ultrasound unit having an ultrasound probe, implemented on a portable computing device. The navigator includes a trained orientation neural network to receive a non-canonical image of a body part from the mobile ultrasound unit and to generate a transformation associated with the non-canonical image, the transformation transforming from a position and rotation associated with a canonical image to a position and rotation associated with the non-canonical image; and a result converter to convert the transformation into orientation instructions for a user of the probe and to provide and display the orientation instructions to the user to change the position and rotation of the probe.

GUIDED DIALOGUE USING LANGUAGE GENERATION NEURAL NETWORKS AND SEARCH

Absstract of: AU2023346892A1

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for enabling a user to conduct a dialogue. Implementations of the system learn when to rely on supporting evidence, obtained from an external search system via a search system interface, and are also able to generate replies for the user that align with the preferences of a previously trained response selection neural network. Implementations of the system can also use a previously trained rule violation detection neural network to generate replies that take account of previously learnt rules.

DATA-FREE POST-TRAINING QUANTIZATION METHOD AND APPARATUS, DEVICE, AND STORAGE MEDIUM

Absstract of: WO2025055499A1

The present application relates to the technical field of neural networks, and provides a data-free post-training quantization method and apparatus, a device, and a storage medium. The method comprises: acquiring data distribution inputted by a user, wherein the data distribution is distribution to which an input activation value of each network layer of a floating-point model conforms; inputting random data into the floating-point model to obtain the input activation value of each network layer; performing statistical analysis on the input activation value of each network layer on the basis of the data distribution to obtain a data range of the input activation value of each network layer; determining a quantization parameter of the input activation value of each network layer on the basis of endpoint values of the data range; and in the process of using the floating-point model for reasoning, using the quantization parameter of the input activation value of each network layer to perform quantization processing on the input activation value generated in the reasoning process of each network layer. By means of the solution of the present application, input activation values can be quantized without data.

SYSTEMS, METHODS, AND APPARATUSES FOR IMPLEMENTING TRANSPARENT MODELS FOR COMPUTER VISION AND IMAGE RECOGNITION UTILIZING DEEP LEARNING NON-TRANSPARENT BLACK BOX MODELS

Absstract of: US2025095347A1

Described herein are means for systematically generating transparent models for computer vision and image recognition utilizing deep learning non-transparent black box models. According to a particular embodiment, there is a specially configured system for generating an explainable AI model by performing operations, including: training a Convolutional Neural Network (CNN) to classify objects; training a Convolutional Neural Network (CNN) to classify objects from training data having a set of training images; training a multi-layer perceptron (MLP) to recognize both the objects and parts of the objects; generating the explainable AI model based on the training of the MLP; receiving an image having an object embedded therein, wherein the image forms no portion of the training data for the explainable AI model; executing the CNN and the explainable AI model within an image recognition system, and generating a prediction of the object in the image via the explainable AI model; recognizing parts of the object; providing the parts recognized within the object as evidence for the prediction of the object; and generating a description of why the image recognition system predicted the object in the image based on the evidence comprising the recognized parts.

DEEP NEURAL NETWORK WITH MULTIPLE LAYERS FORMED OF MULTI-TERMINAL LOGIC GATES

Absstract of: CN119183567A

A deep neural network circuit having a plurality of layers formed by a multi-terminal logic gate is provided. In one aspect, a neural network circuit includes a plurality of logic gates disposed into a plurality of layers and a plurality of logic connectors disposed between each pair of adjacent layers. Each of the logic connectors connects an output of the first logic gate to an input of the second logic gate, and each of the logic connectors has one of a plurality of different logic connector states. The neural network circuitry is configured to be trained to implement a function by finding a set of logical connector states of logical connectors that cause the neural network circuitry to implement the function.

ARTIFICAL NEURAL NETWORK MODEL LEARNING APPARATUS AND INFERENCE SYSTEM FOR VOICE AND SOUND RECOGNITION

Absstract of: WO2025053464A1

The present invention relates to an artificial neural network model learning apparatus and inference system for speech and sound recognition, which comprise: a first learning agent that trains a first artificial neural network model for inference by using pieces of first audio data among a plurality of pieces of audio data collected from a first inference terminal; and a first evaluation agent that evaluates the first artificial neural network model trained by the first learning agent by using pieces of second audio data among the plurality of pieces of audio data collected from the first inference terminal and distributes, through a local network, the first artificial neural network model trained by the first inference terminal according to an evaluation result.

SYSTEM AND METHOD OF RECOGNIZING KEYWORDS IN AN AUDIO STREAM

Absstract of: WO2025052386A1

A system and method of recognizing at least one keyword in an audio stream may include receiving at least one textual representation of at least one respective target keyword of interest. For at least one textual representation, applying a first pretrained Neural Network (NN) model on the textual representation, to obtain a parameter vector representing the respective target keyword, and utilizing the parameter vector as weights in a second NN model. Receiving a target audio stream, and applying the second NN model on the target audio stream, to obtain an utterance indication representing utterance of the respective target keyword within the target audio stream.

SYSTEMS AND METHODS FOR MACHINE LEARNED IMAGE CONVERSION

Absstract of: US2025086753A1

A computer system is provided for converting images through use of a trained neural network. A source image is divided into blocks and context data is added to each pixel block. The context blocks are split into channels and each channel from the same context block is added to the same activation matrix. The action matrix is then executed against a trained neural network to produce a changed activation matrix. The changed activation matrix is then used to generate a converted image.

HYBRID EXPLAINABLE ARTIFICIAL INTELLIGENCE SYSTEM

Absstract of: US2025086447A1

A hybrid explainable artificial intelligence system may include a shallow learning model and a deep learning model. The shallow learning model may be a machine learning system. The deep learning model may be a neural network. The system may input a data set into both the shallow learning model and the deep learning model. Both the shallow learning model and the deep learning model may produce an output. When there is a common output between the shallow learning model and the deep learning model, the process performed by the shallow learning model may be used to formulate an explanation of the process performed by the deep learning model. The explanation of the process performed by the deep learning model may be used to raise the sensitivity of one or more components of the data set. Such components may include a word or phrase within a transcript.

Singing Voice Separation with Deep U-Net Convolutional Networks

Absstract of: US2025087232A1

A system, method and computer product for training a neural network system. The method comprises applying an audio signal to the neural network system, the audio signal including a vocal component and a non-vocal component. The method also comprises comparing an output of the neural network system to a target signal, and adjusting at least one parameter of the neural network system to reduce a result of the comparing, for training the neural network system to estimate one of the vocal component and the non-vocal component. In one example embodiment, the system comprises a U-Net architecture. After training, the system can estimate vocal or instrumental components of an audio signal, depending on which type of component the system is trained to estimate.

ARTIFICIAL NEURAL NETWORK MODEL LEARNING APPARATUS AND INFERENCE SYSTEM FOR SPEECH AND SOUND RECOGNITION

Absstract of: WO2025053464A1

The present invention relates to an artificial neural network model learning apparatus and inference system for speech and sound recognition, which comprise: a first learning agent that trains a first artificial neural network model for inference by using pieces of first audio data among a plurality of pieces of audio data collected from a first inference terminal; and a first evaluation agent that evaluates the first artificial neural network model trained by the first learning agent by using pieces of second audio data among the plurality of pieces of audio data collected from the first inference terminal and distributes, through a local network, the first artificial neural network model trained by the first inference terminal according to an evaluation result.

Method for controlling an electronic device using a machine learning model

Absstract of: GB2633414A

A computer-implemented method involves processing a received input data item using an epistemic neural network of a trained machine learning model. A plurality of possible beliefs from the input data item are derived, e.g. using the epistemic neural network and a translator module. A probability associated with each possible belief is calculated, preferably using a reasoning module to generate a trigger graph. A certainty associated with the probability for each possible belief is calculated, preferably using a d-NNF module. In some embodiments, activation of each output neuron j in the epistemic neural network is the parameter aj of a K-dimensional Dirichlet distribution. In some embodiments, one of the possible beliefs is selected based on the calculated probabilities and certainties, e.g. by selecting a belief with probability and certainty values greater than threshold values. An electronic device may be controlled based on the selected belief.

SYSTEM, METHOD, AND COMPUTER PROGRAM PRODUCT FOR SAVING MEMORY DURING TRAINING OF KNOWLEDGE GRAPH NEURAL NETWORKS

Absstract of: WO2023215214A1

Systems, methods, and computer program products are provided for saving memory during training of knowledge graph neural networks. The method includes receiving a training dataset including a first set of knowledge graph embeddings associated with a plurality of entities for a first layer of a knowledge graph, inputting the training dataset into a knowledge graph neural network to generate at least one further set of knowledge graph embeddings associated with the plurality of entities for at least one further layer of the knowledge graph, quantizing the at least one further set of knowledge graph embeddings to provide at least one set of quantized knowledge graph embeddings, storing the at least one set of quantized knowledge graph embeddings in a memory, and dequantizing the at least one set of quantized knowledge graph embeddings to provide at least one set of dequantized knowledge graph embeddings.

SYSTEM, METHOD, AND COMPUTER PROGRAM PRODUCT FOR ACTIVE LEARNING IN GRAPH NEURAL NETWORKS THROUGH HYBRID UNCERTAINTY REDUCTION

Absstract of: CN119156615A

Systems, methods, and computer program products may (i) obtain a graph comprising a plurality of edges and a plurality of nodes for the plurality of edges, each tagged node in a subset of the tagged nodes being associated with a tag, and each untagged node in a subset of the untagged nodes being not associated with a tag; (ii) training a graph neural network (GNN) using the graph and the label for each marked node, wherein training the GNN generates a prediction for each node; (iii) generating a candidate pool of candidate nodes from the subset of unlabeled nodes; (iv) generating a predictive tag for each candidate node using a tag propagation algorithm; (v) selecting a candidate node in the candidate pool of candidate nodes associated with a maximum mixed entropy reduction of the graph; and (vi) providing the selected candidate nodes for marking.

SYSTEM AND METHOD FOR MENTAL HEALTH DISORDER DETECTION SYSTEM BASED ON WEARABLE SENSORS AND ARTIFICIAL NEURAL NETWORKS

Absstract of: US2025078998A1

According to various embodiments, a machine-learning based system for mental health disorder identification and monitoring is disclosed. The system includes one or more processors configured to interact with a plurality of wearable medical sensors (WMSs). The processors are configured to receive physiological data from the WMSs. The processors are further configured to train at least one neural network based on raw physiological data augmented with synthetic data and subjected to a grow-and-prune paradigm to generate at least one mental health disorder inference model. The processors are also configured to output a mental health disorder-based decision by inputting the received physiological data into the generated mental health disorder inference model.

VIDEO RETRIEVAL BASED CONTEXTUALIZED LEARNING

Absstract of: US2025077859A1

Methods are provided for generating distilled multimedia data sets tailored to user's persona and/or task(s) to be performed associated with an enterprise network and enable interactive contextual learning using a multi-modal knowledge graph. Methods involve obtaining multimedia data from one or more data sources related to operation or configuration of an enterprise network and determining context for generating a distilled multimedia data set based on at least one of user input and user persona. The methods further involve generating, based on the context, the distilled multimedia data set that includes a set of multimedia slices generated from the multimedia data using a multi-modal knowledge graph. The multi-modal knowledge graph is generated using a graph neural network and indicates relationships among a plurality of slices of the multimedia data. The methods further involve providing the distilled multimedia data set for performing one or more actions associated with the enterprise network.

METHOD AND DEVICE FOR ENABLING AN EXPLANATION AND/OR VERIFICATION OF A BEHAVIOUR OF A NEURAL NETWORK TRAINED WITH A TRAINING DATABASE, AND VEHICLE WITH AI

Absstract of: WO2025045319A1

The invention relates to a method for explaining and/or verifying a behaviour of a neural network trained with a training database, said method having the following steps: - determining sensor information by means of a sensor, so that sensor information is available, - applying the neural network to the sensor information, so that target information is available, - determining similarity information, wherein a comparison is made of the sensor information or part of the sensor information with an information database, so that the similarity information is available, linking the similarity information with the target information, so that a target information tuple is available, on the basis of which the explanation and/or verification of the behaviour of the neural network can be realised.

IDENTIFYING TRANSFORMATIONS IN DATA FABRIC USING COUNTERFACTUAL EXPLANATIONS OF ENTITY MATCHING

Absstract of: US2025077877A1

A computer-implemented method for improving entity matching in a probabilistic matching engine can train a graph neural network (GNN) model on an output of a probabilistic matching engine to perform entity matching and determine counterfactual explanations for non-matches of entities. A list of data transformations can be identified by actionable recourse using the GNN model. The list of data transformations can be ranked, using the GNN model, based on computational overhead and an estimated improvement in entity matching within the probabilistic matching engine.

MACHINE LEARNING METHODS TO PRODUCE GENERAL STRUCTURED PREDICTIONS FROM GLOBAL EVENT DATA FOR THE PAST, PRESENT AND FUTURE

Absstract of: US2025077898A1

A computer-implemented method comprising a mathematical formulation of the social energy flow of Karma and a machine learning framework combining mathematical graph computation, which form a global sentiment aggregator for the measurement of social energy flows and which is formulated as a data processing, filtering and sampling framework which fuses structural, interpretable graph machine learning with graph neural networks and introduces a Graph Attention Mechanism (GAM) whereby machine learning guides an interpretable graph computation substrate in order to generate general, structured predictions for the past, present and future. This enables the framework to handle large data sets on a global scale in the graph neural network while affording interpretability within the less scalable graph computational substrate which in turn optimizes the use of memory, computer storage and processing power over traditional designs, thereby making global-scale computations on commodity hardware feasible.

ELECTRONIC DEVICE, METHOD, AND NON-TRANSITORY COMPUTER-READABLE STORAGE MEDIUM FOR IDENTIFYING SPEECH SIGNAL CORRESPONDING TO DESIGNATED TEXT

Absstract of: WO2025048121A1

A processor of an electronic device according to an embodiment may be configured to acquire, from a first neural network to which a speech signal received via a microphone has been input, a first sequence of feature information about portions of the speech signal corresponding to designated frame units. The processor may be configured to acquire, from a second neural network to which the designated text has been input, a second sequence of one or more phonetic symbols for the designated text. The processor may be configured to acquire, from a third neural network to which the first sequence and the second sequence have been input, a first parameter indicating the degree of correspondence between the speech signal and the designated text. The processor may be configured to acquire one or more second parameters indicating the degree of correspondence between the portions of the speech signal and each of the one or more phonetic symbols included in the second sequence.

ELECTRONIC DEVICE, METHOD, AND COMPUTER-READABLE STORAGE MEDIUM FOR IDENTIFYING SPEECH SIGNAL CORRESPONDING TO DESIGNATED TEXT

Absstract of: WO2025047998A1

A processor of an electronic device according to an embodiment may be configured to acquire, from a first neural network to which a speech signal received via a microphone has been input, a first sequence of feature information about portions, corresponding to designated frame units, of the speech signal. The processor may be configured to acquire, from a second neural network to which a designated text has been input, a second sequence of one or more phonetic symbols for the designated text. The processor may be configured to acquire, from a third neural network to which the first sequence and the second sequence have been input, a first parameter indicating the degree of correspondence between the voice signal and the designated text. The processor may be configured to acquire one or more second parameters indicating the degree of correspondence the portions of the voice signal have with respect to each of the one or more phonetic symbols included in the second sequence.

METHODS AND SYSTEMS FOR THE DESIGN AND USE OF INTELLIGENT LAYER SHARED NEURAL NETWORKS AND MODEL ARCHITECTURE SEARCH ALGORITHMS

Absstract of: WO2025049978A1

Presented herein are systems and methods for the use and/or automated design of neural networks (NNs) with layer shared architecture (e.g., an intelligent layer shared (ILASH) neural architecture). In certain embodiments, the NN with layer shared architecture comprises a base set of layers (e.g., base layer shared model) and one or more branches extending from the base set of layers. Each branch may include one or more layers from the base set of layers and one or more additional layers different from the base set of layers, each branch designed and trained to perform a particular unique task on a common set of input data. As a result, the NN will share some layers among multiple tasks. Moreover, presented herein are techniques for using a predictive neural network search algorithm to create the branched network of the layer shared architecture.

END-TO-END TEXT-TO-SPEECH CONVERSION

Nº publicación: US2025078809A1 06/03/2025

Applicant:

GOOGLE LLC [US]
Google LLC