43 results about "Sensory input" patented technology

The invention relates to the technical field of neuromuscular electrical stimulation, and provides a multi-modal feedback self-adaptive adjustment neuromuscular regulation and control treatment system, which comprises wearable working equipment for outputting pulse electric signals to adjust nerves and muscles to form movement or adjustment feeling, electromyographic signals, motion data and user feeling input are collected to conduct feedback adjustment on output parameters of the output pulse electric signals; the handheld control equipment is used for being connected with the wearable working equipment and controlling the output working mode and the working state of the wearable working equipment; and the server-side equipment is used for managing user information and preference of the handheld control equipment, setting function authority of the handheld control equipment, acquiring data acquired by the wearable working equipment, and realizing self-adaptive output of output parameters of the pulse electric signals of the wearable working equipment based on a self-adaptive reinforcement learning algorithm. Adaptive optimization of millisecond-level dynamic electrical stimulation parameters is realized through a multi-modal feedback data fusion technology in combination with a layered reinforcement learning architecture.

The invention discloses an energy storage scheduling method and system in combination with dynamic load prediction, and particularly relates to the field of risk prediction of energy storage scheduling, comprising the steps of constructing a fusion structure comprising a perception input set, a disturbance trend set and a system variable set, and forming a perception input structural body in a scheduling process; comparing the perception input structural body with the behavior track sample set, generating a residual structural body and marking an abnormal behavior candidate section; and processing the abnormal behavior candidate section, and generating an abnormal confirmation result through a freezing mechanism and a trend consistency judgment process. An abnormal behavior candidate section caused by false load change is identified by constructing a sensing input structural body and introducing a behavior trajectory comparison mechanism, so that a misjudgment instruction is frozen before response, trend consistency verification is implemented, frequent start and stop and response drift of an energy storage system caused by sensing errors are avoided, and the fault-tolerant capability of the system is improved.

This invention presents a computer-implemented method for answering polls and surveys using large language models (LLMs), a novel approach that leverages an LLM's ability to emulate a group of human responses for diverse data collection. The system involves training a single or multi-modal LLM on a comprehensive dataset comprising one or more categories of text, image, sound and other sensory input data, which can be continuously updated with current events and trends, to ensure accurate representation of human behaviors and opinions. Utilizing a user interface, the method includes receiving survey and poll questions, posing these questions to a unique instance of the trained LLM and receiving the answers, recording the question-answer sessions, and compiling the results for presentation. Additionally, the system prioritizes data privacy and integrates a feedback mechanism for continuous improvement, providing an efficient solution to answering surveys or polls in the digital age.

In various examples, a technique for performing a task includes converting one or more sensory inputs obtained using one or more sensors of a machine into a plurality of segments. The technique also includes, for each segment included in the plurality of segments, generating, via execution of a machine learning model, a caption for the segment, and storing, in a data store, a representation of the caption in association with the segment. The technique further includes performing, by the machine, one or more actions based at least on one or more queries of the data store.

An emotion-aware computing operating system integrates multimodal sensory inputs, stratified memory, cyclehit scoring, and history score values to predict user tasks, orchestrate autonomous cognitive agents, and adapt interface outputs in real time. A prediction module applies cyclehit scoring derived from historical task cycles and weighted history score values to forecast workflows.An orchestration kernel selects and coordinates agents, redistributes subtasks, and negotiates dynamically under cognitive load. A stratified memory fabric maintains ephemeral, situational, and long-term user models for personalization. An adaptive interface layer adjusts informational density and tool availability based on inferred user state. A certification and licensing API enforces agent onboarding, compliance, and monetization policies, requiring registration of performance metrics prior to integration. Embodiments include software, cloud, edge, and robotic platforms, enabling monetizable deployment across healthcare, finance, education, enterprise, and ambient device ecosystems.

A multifunctional sensory device designed to aid individuals with sensory processing disorders (SPD) in achieving self-regulation is disclosed. The device can be in the form of a plush animal or a backpack. The device includes a plurality of removable weights to provide deep pressure, a scent emitter for calming or alerting smells, a Bluetooth speaker for producing calming sounds, and an embedded vibration coil for adjustable tactile and heating stimulation. The device also features various textured surfaces and modular components for customizable sensory input. Additionally, a connected system monitors and adjusts sensory inputs based on user feedback.

The invention relates to digital watermarking of a machine learning model. In one embodiment, a method is disclosed that includes detecting, by a computing device, a sensory input; identifying, using a machine learning model, one or more attributes associated with the machine learning model, wherein the attributes are identified based on sensory inputs according to training of the model; and presenting the attribute as an output. The identification may be performed at least in part by an inference engine interacting with the model. The sensory input may include an input image received from a camera, and the model may identify an attribute based on an input object in the input image according to training of the model. The model may include a convolutional neural network trained using training data that associates training sensory inputs with attributes. The training sensory input may include a training image having a training object, and the input object may be classified in the same category as the training object.

Implementations of a fully configurable, modularized home robot are described that implement trained neural network classifiers to solve problems of providing home health care and monitoring of the elderly and / or infirm, providing entertainment for the family, providing environmental and safety monitoring coupled with mechanisms to clean the air and remedy indoor climates such as humidity and temperature, and provide a mechanism for caring for pets left at home when the owner is away based upon sensory input.

The processing subsystem communicates with brain regions of an animal subject that are responsible for forming concepts, without sensory input. The processing subsystem receives brain signals representing concepts formed by these brain regions, without sensory input. The processing subsystem processes the received brain signals to convert the concepts into computer-readable data that represents their tangible form. The processing subsystem identifies data records associated with one or more data elements of the computer-readable data, and also associated with real-world objects associated with the concepts.

The invention discloses a self-service terminal based on multi-sensory interaction, and the terminal comprises a multi-sensory input module which is used for receiving voice, gesture and AR / VR input; the instruction analysis engine is used for analyzing the input information into a unified operation instruction; the task distribution module is used for distributing tasks to the corresponding processing units; the mode self-adaption module is used for automatically switching interaction modes; the database module is used for storing related data and examples; the voice synthesis and special effect module is used for carrying out vivid voice broadcast; the gesture auxiliary learning module is used for assisting gesture learning through AR / VR; the gesture auxiliary learning module can adjust a learning guiding strategy and prompting frequency in real time according to the gesture execution condition of the user, and it is ensured that the user effectively learns gesture operation; according to the method, more flexible and convenient operation selection can be provided for the user by integrating multiple interaction modes such as voice, gestures, AR / VR and the like, and the user operation threshold is greatly reduced.

The invention belongs to the technical field of upper limb auxiliary training, and discloses an upper limb auxiliary training device based on the early stage of hemiplegia, which comprises an analysis module, a design module, a training module and an optimization module, according to the training device, through the integrated circuit and the pressure sensor, feelings are input and transmitted to a patient in real time, and stimulation intensity and auxiliary strength are adjusted according to feedback of the patient. The design objective is to provide upper limb extensor active movement induction training for early hemiplegic patients, and operation repeatability and safety are both considered. Clinical practice verifies that the active motion induction training of the upper limb extensor in the early stage of hemiplegia has remarkable effects in promoting motion recovery and improving the functions of the upper limbs. A new technical solution is provided for early hemiplegia rehabilitation treatment, the workload of primary therapists is relieved, and the development of early rehabilitation is promoted. The method is of great significance in promoting functional recovery of hemiplegic patients, improving the life quality of the patients and relieving family and social burdens.

The invention discloses a digital integrated brain injury nerve regulation diagnosis and treatment platform, and relates to the technical field of brain injury diagnosis and treatment, and the technical scheme is characterized in that the platform comprises a multi-mode brain function monitoring module used for collecting brain function data of a brain injury consciousness disorder patient in real time; the multi-sensory nerve regulation and control stimulation module is used for cooperatively applying multi-sensory stimulation to the patient; and the main control and data processing module receives and processes the multi-modal signals transmitted by the multi-modal brain function monitoring module, generates a personalized nerve regulation scheme by extracting brain function characteristics, and controls the multi-sensory nerve regulation stimulation module to execute the personalized nerve regulation scheme at the same time. According to the method, multi-sensory stimulation is cooperatively applied to the patient, the problem of pain points of traditional distributed non-standard stimulation is solved, the core requirement of the patient for multi-channel sensory input awakening is met, meanwhile, a complete closed loop of monitoring-analysis-intervention-feedback is formed, and dynamic evaluation and accurate intervention on the brain function state of the patient with acute brain injury consciousness disorder are achieved.

A computer implemented method for translating a sensory input to a sensory output for communication between first and second entities. The first entity includes a user and the second entity includes a user or a computing system. The method includes obtaining a first profile specifying a communication capability of the first entity, and a second profile specifying a communication capability of the second entity and obtaining a trained Machine Learning (ML) model operable to map an input sensory communication in accordance with one of the first or second profiles to an output sensory communication in accordance with the other of the profiles. The method further includes receiving an input sensory communication from one of the first or second entities, using the ML model to map the input sensory communication to an output sensory communication, and providing the output sensory communication to the other of the entities.

The present disclosure provides a control system for a humanoid robot comprising a bipedal action model (BAM) with hierarchical architecture including a beta model executing cognitive tasks at lower frequency, ingesting multimodal sensory inputs including visual data and natural language instructions, and an alpha model executing reactive tasks at higher frequency, communicatively coupled to the beta model. The BAM is trained on retargeted robot training data derived from robot-free training data. At runtime, the BAM outputs continuous control commands as parallel-generated action chunks controlling at least 18 degrees of freedom. The system includes a wearable collection apparatus capturing movement data from a human operator without physical connection to the robot, and a retargeting module translating robot-free training data into robot training data by solving embodiment mismatches between human and robot kinematic structures.

A system and method for validating unsupervised machine learning models. The method includes: analyzing, via unsupervised machine learning, a plurality of sensory inputs associated with a machine, wherein the unsupervised machine learning outputs at least one normal behavior pattern of the machine; generating, based on the at least one normal behavior pattern, at least one artificial anomaly, wherein each artificial anomaly deviates from the at least one normal behavior pattern; injecting the at least one artificial anomaly into the plurality of sensory inputs to create an artificial dataset; and analyzing the artificial dataset to determine whether a candidate model is a valid representation of operation of the machine, wherein analyzing the artificial dataset further comprises running the candidate model using the artificial dataset as an input.

In various examples, a technique for performing end-to-end navigation using a generative world model includes converting a set of sensory inputs received by a machine at a current time step into a set of embedded features. The technique also includes generating, via execution of one or more neural networks, one or more states associated with the current time step based at least on the set of embedded features, a history of states preceding the current time step, and a first set of actions associated with a previous time step. The technique further includes converting, via execution of the one or more neural networks, the one or more states into a set of predictions associated with the current time step, and performing, by the machine, a second set of actions associated with the current time step based on the set of predictions.

A system and method for unsupervised prediction of machine failures. The method includes monitoring sensory inputs related to at least one machine; analyzing, via at least unsupervised machine learning, the monitored sensory inputs, wherein the output of the unsupervised machine learning includes at least one indicator; identifying, based on the at least one indicator, at least one pattern; and determining, based on the at least one pattern and the monitored sensory inputs, at least one machine failure prediction.

A training system may allow for intelligent training of an athlete by utilizing an array of sensors positioned proximal an athlete and allow for the generation of a feedback strategy. The feedback strategy may select at least one manner of communicating training information to the athlete from a plurality of different available manners of communicating training information. An athletic potential of the athlete may be computed in response to athletic behavior of the athlete detected by the array of sensors. The potential module can further estimate a current skill level of the athlete in response to the sensed athletic behavior of the athlete. A difference between the athletic potential of the athlete and the current skill level may be determined the computing device, which allows for the intelligent assignment of training drills selected to increase a proficiency in at least one athletic task associated with the current skill level.

A method for training a style encoder of a neural network. Sensory input variables, which represent a movement of a system and surroundings of the system, are compressed to an abstract driving situation representation in at least one portion of a latent space of the neural network, using a trained situation encoder of the neural network. The sensory input variables are compressed to a driving style representation in at least one portion of the latent space, using the untrained style encoder. The driving style representation and the driving situation representation are decompressed from the latent space to output variables, using a style decoder of the neural network. A structure of the style encoder is changed to train the style encoder until the output variables of the style decoder represent the movement.

The invention discloses long-term perception for robotic systems and applications. In various examples, techniques for performing a task include converting one or more sensing inputs obtained using one or more sensors of a machine into a plurality of segments. The technique further includes, for each segment included in the plurality of segments, generating a description for the segment via execution of a machine learning model; and storing the representation of the description in a data store in association with the segment. The technique further includes performing, by the machine, one or more actions based at least on the one or more queries on the data store.

Methods of sensory input integrity attestation are provided. Artifacts included within devices under test inject a known noise signal into the output signal of one or more output devices that are detectable by one or more input devices (i.e., sensors) of an embedded device, and monitor the received input data. By comparing the received signal against the expected noise signal, attestation of the validity of sensory input data is possible. Such sensory input data attestation is capable either locally or using a remote attestation device with knowledge of the expected data stream.

The invention is suitable for the technical field of power grid intelligent control, and provides an intelligent control method and device for a low-carbon park optical storage direct-flexible system based on brain emotion learning, and the method comprises the steps: obtaining the output voltage and output current of each distributed power generation unit; calculating a global average voltage and a weighted average current; for each distributed power generation unit, generating a sensory input signal and an emotion signal; inputting the sensory input signal and the emotion signal into a brain emotion learning model to generate a compensation signal; superposing the compensation signal to a local droop control link, and correcting the reference voltage; wherein the sensory input signal represents the voltage deviation and the current distribution deviation of the system, and the emotion signal represents the instantaneous value of the voltage deviation of the system, the integral of the voltage deviation and the current distribution deviation and the self-feedback of the compensation signal. According to the method, voltage stabilization and current accurate distribution are realized through hierarchical cooperative control, the problem of controller performance deterioration caused by model mismatch is solved, and the robustness in a high-uncertainty scene is improved.

Applying therapeutic neural stimuli involves monitoring for at least one of sensory input and movement of a user. In response to detection of sensory input or user movement, an increased stimulus dosage is delivered within a period of time corresponding to a duration of time for which the detected sensory input or user movement gives rise to masking, the increased stimulus dosage being configured to give rise to increased neural recruitment.

Systems, apparatus and methods implemented in algorithms, hardware, software, firmware, logic, or circuitry may be configured to process data and sensory input to determine whether an object external to an autonomous vehicle (e.g., another vehicle, a pedestrian, road debris, a bicyclist, etc.) may be a potential collision threat to the autonomous vehicle. The autonomous vehicle may be configured to implement interior active safety systems to protect passengers of the autonomous vehicle during a collision with an object or during evasive maneuvers by the autonomous vehicle, for example. The interior active safety systems may be configured to provide passengers with notice of an impending collision and / or emergency maneuvers by the vehicle by tensioning seat belts prior to executing an evasive maneuver and / or prior to a predicted point of collision.

A method including: capturing or receiving data associated with behavior of a user of a virtual reality device during a session in a virtual environment, wherein the data includes sensory inputs associated with the user from one or more sensors and information associated with user parameters; initiating authentication of the user in the virtual environment; comparing the captured or received data with historic data of the user, wherein the historic data is associated with behavior of the user monitored for a plurality of sessions over a period of time in the virtual environment; determining a score based on the comparison; comparing the score with a predefined threshold score; and in response to determining that the score is above the predefined threshold score, authenticating the user.

A processing subsystem communicates with a region of an animal subject's brain that is responsible for forming concepts without sensory input. The processing subsystem receives brain signals that are representative of a concept formed by the region of the brain without sensory input. The processing subsystem processes the received brain signals to convert the concept to computer readable data that is representative of a tangible form of the concept. The processing subsystem identifies a data record that is associated with one or more data elements of the computer readable data and that is also associated with a real-world object that is related to the concept.

Exemplary embodiments include systems and methods for pairing a user to an available parking space, the system comprising: a positional device; at least one sensory input device positioned on an infrastructure point located proximally to at least one parking space; and a central server platform comprising a server having a processor and a memory unit communicatively coupled with the processor, the memory unit storing instructions which, when executed, perform a method. The method comprises: receiving a first signal from the positional device, the first signals indicating an intention of a user to seek parking and further indicating a location of the user; receiving a second signal from the at least one sensory input device, the second signal indicating availability of a parking space; determining, from the second signal, which of the parking spaces are available to the user; and pairing the user with a parking space.

An isolation therapy device worn on the head for reducing sensory input and delivering therapeutic treatment for neurological and psychological conditions, including anxiety, PTSD, autism, cognitive decline, and concussion. The device comprises a multi-layer construction including: an innermost comfort layer, an intermediate adjustable compression or suspension layer, and an outer fabric shell layer. The device further includes optional modular systems for sensory reduction and therapeutic enhancement, including modules for visual and auditory isolation, aromatherapy, heating, cooling, vibration, and EEG monitoring. Modular components may be removable, adjustable, and customizable. The device optionally integrates smart sensors for biometric monitoring and real-time feedback. A digital application may be used in conjunction with the device to enable control, tracking, and personalization of therapy parameters. The modular nature allows for individualized treatment protocols and adaptability across a range of conditions.

The Fidget Knob is a compact, handheld device designed to provide tactile stimulation and promote focus and relaxation. Consisting of two rotating knobs, possibly with textured surfaces, it offers a satisfying tactile experience for users to fidget with during moments of stress or boredom. The Fidget Knob's ergonomic design allows for easy manipulation and discreet use in various settings, such as classrooms, offices, meetings, and public spaces. The Fidget Knob's reasonably silent behavior allows it to be used & felt by the user without distracting nearby people. Research suggests that fidgeting can help improve concentration and reduce stress levels by providing sensory input and promoting a sense of calm. The Fidget Knob offers a convenient and socially acceptable way for individuals to engage in fidgeting behavior, potentially enhancing cognitive performance and well-being.

A method for determining driving comfort in a motor vehicle is provided, in which an AI is used to determine driving comfort by evaluating sensor data from the vehicle's sensors, where the sensor data includes optical signals. Since not only the interior climate but also, or alternatively, optically detectable influences occurring only during driving, especially vibrations, are considered as effects on subjectively perceived driving comfort, a high level of driving comfort is achieved. By coupling this with acoustic and mechanical sensors, driving comfort can be evaluated using all sensory input.