Abstract: A system described herein may provide for the adaptation and/or expansion of a natural language processing (“NLP”) platform, that supports only a limited quantity of intents, such that the described system may support an unlimited (or nearly unlimited) quantity of intents. For example, a hierarchical structure of agents may be used, where each agent includes multiple intents. A top-level (e.g., master) agent may handle initial user interactions, and may indicate a next-level agent to handle subsequent interactions.

Abstract: A system and method for reducing laser communication terminal pointing uncertainty. The method trains an artificial neural network (ANN) with input data characterizing terminal pointing error and dependent parameters. The method inputs the trained ANN a set of data of these dependent parameters with unknown pointing error. The method uses the ANN output to apply corrections to the terminal pointing solution to reduce pointing uncertainty. The method can condition the ANN generated corrections to avoid cases where application of the ANN correction could exceed the original pointing uncertainty. This conditioning includes computing the Euclidean distance between current ANN input parameter values and values in the ANN training dataset, and bounding the allowed magnitude of the ANN pointing correction. The method can train the ANN incrementally during terminal operation for real-time updates or train the ANN offline with gathered data and implement the trained ANN on the terminal for subsequent links.

Abstract: Various implementations include wearable audio devices configured to record a user's voice without recording other ambient acoustic signals, such as others talking nearby. In some particular aspects, a wearable audio device includes: a frame for contacting a head of a user; an electro-acoustic transducer within the frame and configured to output audio signals; at least one microphone; a voice activity detection (VAD) accelerometer; and a controller coupled with the electro-acoustic transducer, the at least one microphone and the VAD accelerometer, the controller configured in a first mode to: detect that the user is speaking; and record a voice of the user solely with signals from the VAD accelerometer in response to detecting that the user is speaking.

Abstract: A non-transitory computer-readable recording medium stores a learning program that causes a computer to execute a process including: extracting, from a plurality of data groups, a plurality of first data groups having an order; generating, for each data element corresponding to each of the first data groups, an ordered data matrix in which data elements having same order have value corresponding to relationship among the data elements and data elements having different orders have values corresponding to the different orders; and obtaining input tensor data by performing tensor decomposition with the ordered data matrix, inputting the input tensor data to a neural network at time of performing deep machine learning, performing deep machine learning of the neural network, and learning about method for the tensor decomposition.

Abstract: Embodiments may generally be directed components and techniques to detect a request to provide banking account information over a one or more voice connections, identify the requested banking account information, and generate speech data representing the banking account information requested. In embodiments further include communicating the speech data to another device.

Abstract: At least some embodiments are directed to an entity classification system receives informational data associated with an entity. The informational data includes sentences associated with the entity. The entity classification system utilizes a first machine learning model to determine a first contextual meaning among words of a sentence associated with the entity based on a first word embedding technique, and determines at least one category associated with the entity based at least in part on the first contextual meaning. The entity classification system utilizes a second machine learning model to determine a second contextual meaning shared by a set of sentences based on a second embedding technique, and determines a subcategory of the category associated with the entity based at least in part on the second contextual meaning. The entity classification system generates an output including the category and subcategory associated with the entity.

Abstract: In-ear sound pressure level, SPL, is determined that is caused by output audio being converted into sound by a headset worn by a user. The in-ear SPL is converted into a sound sample having units that are suitable for evaluating sound noise exposure. These operations are repeated to produce a sequence of sound samples during playback. This sequence of sound samples is then written to a secure database. Access to the database is authorized by the user. Other aspects are also described and claimed.

Abstract: Systems, apparatuses, methods, and computer program products are disclosed for screening data instances based on a target text of a target corpus. A screening device analyzes a target corpus to generate at least two term dictionaries for the target corpus. The screening apparatus, based on a frequency of a term in the target corpus, determines a term weight for the term; for each data instance, determines term scores for the data instance and the target text based on the term weights; filters the data instances based on the term scores, to generate a short list of data instances; determines term similarity scores between each data instance of the short list and target text based on the term weights; and provides a data instance determined to likely correspond to the target text and an indication of the corresponding term similarity score(s). A term is a word or an n-gram.

Abstract: Provided is a system that transmits a notification to a user electronic equipment associated with a user, based on a user location information. This notification includes information representative of a sound-based promotional campaign for a product at a location known by the user. The system also transmits a location-specific promotional information, associated with this sound-based promotional campaign and based on a user information, to a merchant electronic apparatus associated to this known location.

Abstract: A system and method for automated customer response system mapping and duplication to provide for a common Interactive Voice Response (“IVR”) system import language or structure, comprising a contact center crawler which connects with an existing automated contact center system, may navigate and map that system by recording and selecting each available option or by interfacing with the code that makes up the IVR itself to generate a model, records the response to each selected option, and feeds the options and responses to a tree creator; a CX model generator which organizes the options and responses into a data map such as a tree or more graph-like or map-like structure, where logical conditions or operations such as loops and recursion may be present; and a coding engine that generates programmatic code representing the data model that is imported into another IVR platform for execution.

Abstract: Methods and systems are provided for gaming headset with enhanced off-screen awareness. An audio system that is used for outputting audio signals to a user may be configured for identifying in audio signals, based on parameters or criteria associated with a user of the system, one or more components in the audio signals; determining characteristics of the one or more components based on perception of the audio signals by the user; and adjusting based on the characteristics of the one or more components, the one or more components and/or one or more other components in the audio signals. The audio system may include circuits for handling the required functions. The circuits may include a detection circuit configured for identify the one or more components, one or more adjustment circuits for applying the adjustments, and a controller circuit for controlling applying the adjustments.

Abstract: The present application relates to a human-machine interaction method and device, a computer apparatus, and a storage medium. The method comprises: measuring the current output volume, and if the output volume is less than a first preset threshold, enabling a voice recognition function; acquiring a user's voice message, and measuring the size of the user's voice volume and responding to a user's voice operation; and if the user's voice volume is greater than a second preset threshold, turning down the output volume, and returning to the step of measuring the current output volume.

Abstract: Disclosed herein, among other things, are systems and methods for a hearing device antenna. One aspect of the present subject matter includes a hearing device configured to be worn in an ear of a wearer to perform wireless communication. The hearing device includes a housing, hearing electronics within the housing, and an inverted F antenna or loop antenna disposed at least partially in the housing and configured for performing 2.4 GHz wireless communication. In various embodiments, at least a portion of the antenna protrudes from an exterior of the housing.

Abstract: The present disclosure generally relates to embodiments for video communication interfaces for automatically adjusting a displayed representation of a field-of-view of a camera in response to detecting a change in a scene.

Abstract: A device may receive user personalized data and user activity data identifying tasks and actions performed by a user, and may perform natural language processing on the user personalized data and the user activity data to generate processed textual data. The device may train machine learning models based on the processed textual data to generate trained machine learning models, and may receive, from a client device, a command identifying a particular task to be performed. The device may process the command and the user activity data, with the trained machine learning models, to determine whether a particular action in the user activity data correlates with the particular task. The device may perform actions when the particular action correlates with the particular task.

Abstract: Once participants participate in a video conference the state of information during that video conference is retained for those participants. At a future time, when those participants again meet in a video conference the state of information that was retained is retrieved and presented for them again in the subsequent video conference based upon their identifiers. Information such as the whiteboard, the chat window, shared files, camera settings, images, background, video recording, etc. are retrieved and automatically presented in the new video conference. Alternatively, a participant may be asked if the state of information from a previous video conference should be retrieved. When a participant joins a conference, the current participants grant permission or not for the current contents to be viewed by the new participant. When a participant leaves the conference, or when they all leave, a triggering condition saves the state.

Abstract: Aspects of the subject disclosure may include, for example, a method for spherical video streaming over a network by a processing system including a processor, the method including prefetching audio data for rendering the spherical video; determining whether there is a single source or a plurality of sources of audio in the audio data; responsive to a determination that there is only one dominant source of the audio in the audio data, prefetching video data in a field of view centered on spatial coordinates of the dominant source of audio; responsive to a determination that there is the plurality of sources of the audio in the audio data, sorting the plurality of sources of the audio according to an estimated sound source energy for each source in the plurality of sources, thereby creating a sorted list of the plurality of sources; and prefetching video data in plural fields of view centered on spatial coordinates of each audio source in the sorted list of the plurality of sources, wherein the prefetching is

Abstract: A sound insulation method and a device based on LiFi optical communication, electronic device, and a storage medium are provided. The sound insulation method based on LiFi optical communication includes: zoning a public area and a private area by a light through LiFi optical communication; collecting one or multiple sound sources; extracting a private voice data of one or multiple sound sources from the sound sources; obtaining a sound source location of a private voice data according to the identified voice data; filtering a private voice data whose sound source location is located in the private area from the sound sources; and performing voice communication according to the filtered sound sources. The invention realizes intelligent sound insulation in audio calls and/or video calls.

Abstract: An electronic apparatus is provided. The electronic apparatus includes a memory configured to store an entity identification model, and a processor configured to control the electronic apparatus to: identify a plurality of entities included in an input sentence input based on the entity identification model, acquire a search result corresponding to the plurality of entities, and correct and provide the input sentence based on the search result based on the plurality of entities not corresponding to the search result, wherein the entity identification model may be acquired by learning through an artificial intelligence algorithm to extract a plurality of sample entities included in each of a plurality of sample dialogues. At least a part of a method for identifying an entity from a dialogue may use an artificial intelligence model learned according to at least one of machine learning, a neural network, deep learning algorithm, or the like.

Abstract: Automated methods and systems provide a gallery view in an online meeting user interface that is changeable by gestures of meeting participants. The gallery view displays video feeds of the meeting participants on their respective participant computers, the video feeds being camera-captured views of each of the meeting participants. Software of a video conferencing system recognizes a specific gesture of a meeting participant, wherein the specific gesture is associated with a predefined directional movement of the meeting participant's video feed within the gallery view during the online meeting. The videoconferencing system electronically and automatically repositions the video feed of the meeting participant who made the specific gesture within the gallery view during the online meeting in accordance with the associated predefined directional movement.