Neural networks

SELF-CLASSIFICATION OF NEURAL NETWORKS

Absstract of: WO2025083107A1

The invention relates to a method for categorizing objects detected by n (n = 2, 3...) artificial neural networks in at least one image.

METHOD AND SYSTEM FOR GENERATING TECHNICAL EXPLANATIONS OF ALARMS IN A SCADA SYSTEM

Absstract of: EP4783068A1

0001 According to an aspect of the present inventive concept there is provided a computer-implemented method (1000) for generating technical explanations of alarms in a supervisory control and data acquisition, SCADA, system for an electrical infrastructure, the method (1000) comprising: providing (1100) a neural network, NN, model trained to detect anomalies in data, wherein the NN model is trained on a training dataset comprising training data based on operational data retrieved by the SCADA system from the electrical infrastructure and comprising data related to a plurality of system parameters of the SCADA system, receiving (1200) run data based on operational data retrieved by the SCADA system from the electrical infrastructure, the run data comprising data related to the plurality of system parameters, detecting (1300), with the NN model, an anomaly in the run data, and generating (1400) a technical explanation of the detected anomaly by means of an interpretability model, wherein the interpretability model applies a model-agnostic interpretable technique to the run data to determine a contribution of the system parameters in the plurality of system parameters to the detection of the anomaly.

A NEURAL NETWORK SYSTEM WITH MULTIPLE INPUTS AND MULTIPLE OUTPUTS

Absstract of: CN121889838A

A method and apparatus for deep learning. A first input and a second input are accessed. A first embedding of the first input is generated using the bound network. A second embedding of the second input is generated using the bound network. The first embed and the second embed are aggregated to generate a combined embed. The transform function is applied to the combinatorial embedding to generate a transformed combinatorial embedding. The combined embedding of transforms is processed using an unbinding network to extract an embedding of a first transform for the first input and an embedding of a second transform for the second input. An inference function is applied to the embedding of the first transform to generate a first output. An inference function is applied to the embedding of the second transform to generate a second output.

PARALLEL COMPUTING SCHEME GENERATION FOR NEURAL NETWORKS

Absstract of: EP4783073A2

0001 Various embodiments relate to determining a parallel computation scheme for a neural network. A device may receive a computation graph and transform the computation graph into a dataflow graph comprising recursive subgraphs. Each recursive subgraph may comprise a tuple of another recursive subgraph and an operator node, or an empty graph. The device may determine a number of partitioning recursions based on a number of parallel computing devices. For each partitioning recursion, the device may determine costs corresponding to operator nodes, determine a processing order of the recursive subgraphs, and process the recursive subgraphs. To process a recursive subgraph, the device may select a partitioning axis for tensor(s) associated with an operator node of the recursive subgraph. The device may output a partitioning scheme comprising partitioning axes for each tensor associated with the operator nodes. Devices, methods, and computer programs are disclosed.

METHOD AND SYSTEM OF IMAGE CLASSIFICATION USING A DEEP NEURAL NETWORK EMBEDDED WITH MULTI-SCALE SPATIAL ATTENTION MECHANISM

Absstract of: US20260212659A1

0000 The present invention provides a method for image classification by incorporating a deep neural network embedded with multiscale spatial attention mechanism (MSSAM). The method according to the present invention comprises various stages: Stage I—Data preparation stage; Stage II—Model training stage; Stage III—Evaluation and Testing stage; Stage IV—Iterative optimization stage. During Data preparation stage, data is collected from a large and diverse dataset of images relevant to specific classification task. During the Model training stage, the model architecture is established, appropriate loss function is selected, an optimizer and an initial learning rate is chosen, the model is trained on training dataset, monitoring validation performance and experiments are performed with hyperparameters. During the Evaluation and Testing stage, model's performance is evaluated on the validation set using metrics and during Iterative optimization stage, the optimization process is iterated and continuously monitored for best results.

AUTOMATED ANEUPLOIDY SCREENING USING ARBITRATED ENSEMBLES

Absstract of: AU2026205299A1

Abstract Systems and methods are provided for fully automated screening for aneuploidy in a human embryo. An image of the embryo is obtained at an associated imager and provided to a neural network to generate a first clinical parameter. A set of at least one parameter representing one of biometric parameters of one of a patient receiving the embryo, an egg utilized to produce the human embryo, a sperm used to create the embryo, a sperm donor who provided sperm used to create the embryo, and an egg donor who provided the egg is retrieved, and a second clinical parameter is generated from the set of at least one parameter at a predictive model. A composite parameter, representing a likelihood of aneuploidy in the embryo, is generated from the first clinical parameter and the second clinical parameter. Abstract ul u l b s t r a c t

TRAINING REINFORCEMENT LEARNING AGENTS TO LEARN FARSIGHTED BEHAVIORS BY PREDICTING IN LATENT SPACE

Absstract of: US20260208355A1

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training an action selection policy neural network used to select an action to be performed by an agent interacting with an environment. In one aspect, a method includes: receiving a latent representation characterizing a current state of the environment; generating a trajectory of latent representations that starts with the received latent representation; for each latent representation in the trajectory: determining a predicted reward; and processing the state latent representation using a value neural network to generate a predicted state value; determining a corresponding target state value for each latent representation in the trajectory; determining, based on the target state values, an update to the current values of the policy neural network parameters; and determining an update to the current values of the value neural network parameters.

AUTOMATIZED DETECTION OF INTESTINAL INFLAMMATION IN CROHN'S DISEASE USING CONVOLUTIONAL NEURAL NETWORK

Absstract of: US20260212495A1

0000 The invention relates to system and methods for predicting and/or diagnosing of IBD from ultrasound images according to one or more of diagnostic signs.

METHODS AND APPARATUSES FOR CHARACTERIZING CHEMICAL SUBSTANCES, MEASURING PHYSICOCHEMICAL PROPERTIES AND GENERATING CONTROL DATA FOR SYNTHESIZING CHEMCIAL SUBSTANCES

Absstract of: US20260212962A1

Provided is a method for generating a digital representation of a chemical substance. This may involve training aspects of neural networks to represent chemical sub-stances. Provided further relates to applications of the digital representation including a computer program product, a database search engine for identifying chemical sub-stances, apparatuses for generating measurement data associated with chemical substances and control data associated with synthesis specifications for chemical substances.

MACHINE LEARNING MODELS FOR BEHAVIOR UNDERSTANDING

Absstract of: US20260212162A1

0000 A method for performing one or more tasks, wherein each of the one or more tasks includes predicting behavior of one or more agents in an environment, the method comprising: obtaining a three-dimensional (3D) input tensor representing behaviors of the one or more agents in the environment across a plurality of time steps; generating an encoded representation of the 3D input tensor by processing the 3D input tensor using an encoder neural network, wherein 3D input tensor comprises a plurality of observed cells and a plurality of masked cells; and processing the encoded representation of the 3D input tensor using a decoder neural network to generate a 4D output tensor.

Machine-Learning Techniques For Monotonic Neural Networks

Absstract of: AU2026205315A1

Abstract In some aspects, a computing system can generate and optimize a neural network for risk assessment. The neural network can be trained to enforce a monotonic relationship between each of the input predictor variables and an output risk indicator. The training of the neural network can involve solving an optimization problem under a monotonic constraint. This constrained optimization problem can be converted to an unconstrained problem by introducing a Lagrangian expression and by introducing a term approximating the monotonic constraint. Additional regularization terms can also be introduced into the optimization problem. The optimized neural network can be used both for accurately determining risk indicators for target entities using predictor variables and determining explanation codes for the predictor variables. Further, the risk indicators can be utilized to control the access by a target entity to an interactive computing environment for accessing services provided by one or more institutions. Abstract ul b s t r a c t u l

SYSTEM AND METHOD FOR PROCESSING ULTRASOUND IMAGES

Absstract of: US20260212999A1

0000 A computer-implemented method provides real-time visual guidance for orienting an ultrasound probe toward a canonical ultrasound view. The method involves receiving successive ultrasound images acquired by the probe during a scanning procedure. A trained neural network generates a spatial mapping from the received images without reliance on external position or motion sensors. This mapping encodes adjustments in probe position and rotation needed to achieve a target spatial pose associated with the canonical view. The spatial mapping is then transformed into visual guidance data. One or more visual guidance elements, derived from this data, are generated and displayed on a graphical user interface to instruct a user on how to manipulate the probe. The visual guidance elements are continuously updated until the canonical ultrasound view is achieved, helping non-experts capture high-quality diagnostic images.

ABSTRACTION LIBRARY TO ENABLE SCALABLE DISTRIBUTED MACHINE LEARNING

Absstract of: US20260212440A1

One embodiment provides for a non-transitory machine readable medium storing instructions which, when executed by one or more processors, cause the one or more processors to perform operations comprising providing an interface to define a neural network using machine-learning domain specific terminology, wherein the interface enables selection of a neural network topology and abstracts low-level communication details of distributed training of the neural network.

FPGA-ORIENTED DSP PLACEMENT OPTIMIZATION METHOD FOR CNN ACCELERATORS

Absstract of: US20260212104A1

An FPGA-oriented DSP placement optimization method for CNN accelerators includes the following steps: S1. DSP path information extraction: converting a designed netlist into a graph representation, and carrying out data-path DSP node identification and data-path DSP graph building; and S2. datapath-driven DSP placement: distributing data-path DSP nodes to specific positions on an FPGA according to an extracted data-path DSP graph. According to the invention, automated extraction and building of data-path DSP graphs are carried out by means of graph neural network (GCN)-based DSP node classification and min-cost flow (MCF) model optimization algorithms, and compact placement and cascade constraint optimization are used in combination, thereby greatly improving the timing performance and computing efficiency of the placement and also significantly improving the clock frequency and throughput. Therefore, the method provides a universal and efficient FPGA placement solution for multiple CNN accelerator architectures.

LIGHT RENDERING METHOD AND APPARATUS, COMPUTER DEVICE, AND COMPUTER-READABLE STORAGE MEDIUM

Absstract of: US20260212595A1

A light rendering method includes: performing light effect detection on a rendered image, to obtain a detection result; in response to the detection result indicating that a first object vertex not meeting a light rendering condition exists in the rendered image, obtaining, along a light path on which the first object vertex is located, light information of a second object vertex that is in the rendered image and that meets the light rendering condition; training a neural network based on the light information of the second object vertex, to obtain a trained neural network; extracting light information of object vertexes including the first object vertex in a target image by using the trained neural network, to obtain the light information of the object vertexes; and performing light rendering on the target image based on the light information of the object vertexes.

METHOD AND APPARATUS FOR COMPRESSING A NEURAL NETWORK, ELECTRONIC DEVICE, AND AIRCRAFT

Absstract of: US20260212183A1

A method and an apparatus for compressing a neural network. Circuitry receives a neural network comprising a set of parameters being in at least one floating-point number format of a first precision. The circuitry also applies at least one mathematical model on the neural network to determine errors introduced by using arithmetic of a second precision lower than the first precision and propagation through network layers of the neural network due to using the arithmetic of the second precision instead of the first precision. The circuitry then applies a solver to an optimization problem formulated based on the errors determined by the at least one mathematical model to determine a set of optimized parameters. Finally, the circuitry compresses the neural network based on the set of optimized parameters to output a compressed neural network.

SELF-DEVELOPING NEURAL NETWORK DESIGN METHOD INSPIRED BY DNA DAMAGE REPAIR MECHANISM, AND STORAGE MEDIUM

Absstract of: WO2026152352A1

A self-developing neural network design method inspired by a DNA damage repair mechanism, and a storage medium. The method comprises: selecting a neural network model and defining hyperparameters thereof; acquiring an image dataset, and dividing the image dataset into a training set, a validation set and a test set, in order to form sequential tasks; on the basis of data of the first task in the training set, training the neural network model; on the basis of data of the first task in the validation set, testing a trained neural network model; performing self-developing growth on the neural network model in the width and depth directions, in order to increase the scale of the neural network model; repeating the above training and testing processes, determining whether it is necessary to further increase the scale of the neural network model, and if the increase is stopped, acquiring the classification accuracy of the model based on task data in the test set; and repeating the above process until the testing of all the task data in the test set is completed. The scale of a neural network model is dynamically adjusted during a training process, and new neurons are grown to adapt to changing tasks and environments, thereby promoting the intelligent development of neural networks in image classification tasks and the practical application thereof.

MODULAR CONCEPT-BASED LANGUAGE PROCESSING WITH TRACE-DRIVEN RESOURCE TIERING AND DYNAMIC CACHING

Absstract of: US20260211817A1

0000 Modular systems and methods for concept-based processing of natural language are provided. An ontology dictionary stores concepts each having a concept identifier, attributes, and typed relationships. Natural language inputs are mapped to concept identifiers, optionally by deriving intermediate units using transforms or statistical analysis and mapping the intermediate units to concepts, or by mapping raw tokens to concepts. A neural network model processes the concept identifiers to generate outputs. During inference and/or training, a trace engine records activation information associated with concepts and stores activation information or derived statistics in a trace data structure. A resource controller uses the trace data structure to reconfigure, during inference for subsequent inputs, a resource allocation policy that allocates frequently used concept data to a faster memory tier and allocates infrequently used concept data to a slower tier or specialized tail handling. Predictive prefetching and relational group caching may be performed.

SYSTEMS, APPARATUSES, AND METHODS FOR DISEASE PREDICTION

Absstract of: WO2026155960A1

A system for training a graph neural network (GNN) for disease prediction includes at least one memory configured to store instructions and at least one processor configured to execute the instructions to cause the system to obtain a plurality of first parameter values of parameters from individuals, determine a causal structure indicative of one or more causal relationships between the plurality of parameters based on the plurality of parameters, and train the at least one GNN with the one or more causal relationships to create a trained GNN, the trained GNN configured to output a disease prediction based on second parameter values, the second parameter values being parameter values of parameters from a patient.

PERFORMING TASKS USING GENERATIVE NEURAL NETWORKS

Absstract of: US20260212167A1

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for performing tasks. One of the methods includes obtaining a sequence of input tokens, where each token is selected from a vocabulary of tokens that includes text tokens and audio tokens, and wherein the sequence of input tokens includes tokens that describe a task to be performed and data for performing the task; generating a sequence of embeddings by embedding each token in the sequence of input tokens in an embedding space; and processing the sequence of embeddings using a language model neural network to generate a sequence of output tokens for the task, where each token is selected from the vocabulary.

LEARNING A GENERATIVE FUNCTION CONFIGURED TO GENERATE A B-REP GIVEN A CONDITIONING SIGNAL REPRESENTING A GEOMETRY

Absstract of: EP4779510A1

The disclosure notably relates to a computer-implemented method of machine-learning, for learning a generative function configured to generate a B-rep given a conditioning signal representing a geometry. The generative function comprises a vertex neural network, an edge neural network, and a face neural network. The edge and face neural network each further comprise a topological model and a geometrical model. Each of the neural network is configured to generate the respective parts of the B-rep (vertices, edges, and faces) respecting the conditioning signal. This provides an improved solution for designing B-reps given a conditioning signal representing a geometry.

MULTI-VIEW CONVOLUTIONAL NEURAL NETWORKS FOR VIDEO PROCESSING

Absstract of: WO2025058768A1

The present disclosure relates to processing video data. Some aspects involve partitioning input video data into two or more clips, each clip comprising a number of T frames, wherein each frame comprises a frame height, a frame width and a frame channel dimension Cin. Each clip is encoded into S encoded representations comprising a code height, a code width, and a code channel dimension, wherein T and S are integers with T ≥ S > 1. Encoding each clip into the S encoded representations may comprise concatenating all T frames of the clip into an input tensor along the frame channel dimension and encoding the input tensor into the S encoded representations using a convolutional neural network (CNN) encoder.

Method and system for cognitive enhancement of artificial intelligence language models

Absstract of: GB2703359A

A system for enhancing artificial intelligence language models includes a Neurotransmitter Simulation Module (NSM) configured to simulate dynamics of multiple neurotransmitters. The system further includes a State Interpreter configured to generate a multi-dimensional cognitive-emotional state vector based on analysing neurotransmitter levels and an Adaptive Parameter Adjustment Module (APAM) configured to adjust language model parameters based on the cognitive-emotional state vector. A language model is configured to generate natural language outputs using the adjusted parameters. A feedback loop mechanism is configured to evaluate quality of the natural language outputs and provide feedback signals to at least one of the NSM or the APAM. The system may further include a Dynamic Neurotransmitter Balancer to maintain balance among the neurotransmitters. The neurotransmitters may include serotonin, dopamine, norepinephrine, acetylcholine, and gamma-aminobutyric acid (GABA). The state interpreter may comprise a neural network trained to map the neurotransmitter levels to generate the multi-dimensional cognitive-emotional state vector. Figure 1

INTELLIGENT TRANSPORTATION SYSTEMS

Absstract of: EP4778450A2

0001 Transportation systems have artificial intelligence including neural networks for recognition and classification of objects and behavior including natural language processing and computer vision systems. The transportation systems involve sets of complex chemical processes, mechanical systems, and interactions with behaviors of operators. System-level interactions and behaviors are classified, predicted and optimized using neural networks and other artificial intelligence systems through selective deployment, as well as hybrids and combinations of the artificial intelligence systems, neural networks, expert systems, cognitive systems, genetic algorithms and deep learning.

NEURAL NETWORK INFERENCE USING A QUANTIZED KEY-VALUE CACHE

Nº publicación: EP4777947A1 22/07/2026

Applicant:

GDM HOLDING LLC [US]
GDM Holding LLC

Absstract of: WO2025068440A1

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for using a Transformer-based neural network to generate output sequences. To generate the output sequences, the Transformer-based neural network is configured to perform quantized inference.