12 results about "Received spoken" patented technology

As part of a dialog session between a user and an automated assistant, implementations can receive a stream of audio data that captures a spoken utterance including an assistant query, determine, based on processing the stream of audio data, a set of assistant outputs that are each predicted to be responsive to the assistant query, process, using large language model (LLM) output(s), the assistant outputs and context of the dialog session to generate a set of modified assistant outputs, and cause given modified assistant output, from among the set of modified assistant outputs, to be provided for presentation to the user in response to the spoken utterance. In some implementations, the LLM output(s) can be generated in an offline manner for subsequent use in an online manner. In additional or alternative implementations, the LLM output(s) can be generated in an online manner when the spoken utterance is received.

A bidirectional sign language communication system (100) with multisensor glove and haptic feedback, comprising: a wearable glove with a variety of sensors configured to detect hand, finger and wrist movements of a user, wherein the sensors include at least curvature sensors, a pinch detector, a strain sensor, a pressure measuring arrangement and an inertial measurement unit; a processing unit that is functionally coupled with the sensors and configured to process sensor signals and derive sign language tokens using a machine learning model; a wireless communication interface configured to transmit the derived tokens to a companion device for playback as text or speech; wherein the companion device is further configured to receive spoken or textual inputs, convert the inputs into coded tactile patterns, and transmit the patterns to the wearable glove; and a distributed haptic feedback subsystem integrated into the wearable glove and configured to present tactile patterns to the user via fingertip actuators, vibration motors, or a tactile matrix, wherein the system (100) enables bidirectional real-time offline communication between a sign language user and a non-sign language user.

Described herein are examples of hardware-based voice-activated artificial intelligence (AI) systems for versatile task execution which include an input interface, a processing module, a control circuit, a communication module, and an audio device. The input interface may include a microphone configured to receive spoken language input provided by a user in the form of an audio signal and a converter configured to convert the captured audio signal to a digital signal. The processing module may include a central processing unit (CPU) and memory, configured to process the received spoken language input by executing a machine learning algorithm or a natural language processing routine stored in the memory. The control circuit may generate a response or action based on the spoken language input. The communication module may be communicatively linked with an external system. The audio device may be configured to provide a voice-based response or audio feedback to the user.

The present invention relates to a method and apparatus for generating a regression-based gloss-free sign language pose using discrete representation that generates a sign language pose on the basis of (gloss-free) spoken sentence data, and the method comprises the steps of: receiving spoken sentence data as an input and transforming the spoken sentence data into a discretized sign language pose token according to spatial and temporal characteristics by using a self-supervised learning model; and automatically recursively generating a sign language pose through a transformer-based encoder-decoder architecture by using the discretized sign language pose token as an input.

A computer-implemented system and method enable real-time healthcare transaction processing based on ambient audio captured during physician-patient conversations. The system includes one or more microphones to receive spoken dialogue and a speech recognition engine to transcribe the audio into text. A natural language processing module analyzes the text to identify clinical intents corresponding to healthcare transactions. The identified intent is standardized into a structured query format and transmitted to an external system via a communications interface. In response, the system receives external data such as insurance coverage status, cost estimates, or authorization requirements. A user interface displays the external response in real time during the clinical encounter, allowing the physician or patient to make informed decisions. Additional modules may include context monitoring logic to suppress non-actionable utterances and a secure storage engine to log queries and responses. The system automates insurance verification and authorization workflows without requiring manual data entry.

A computing device is described that accepts, a telephone call, from another device, initiated by a caller. Prior to establishing a telephone user interface that receives spoken input from the user and outputs spoken audio from the caller, the computing device executes a call screening service that outputs an audio user interface, to the other device and as part of the telephone call. The audio user interface interrogates the caller for additional information including a purpose of the telephone call, which allows the user to have more context of the telephone call before deciding whether to accept the call or hang up. The computing device outputs a graphical user interface associated with telephone call. The graphical user interface includes an indication of the additional information obtained via the audio user interface that interrogates the caller.

An eyeglass translation system includes a wearable frame, an audiovisual input assembly, a microprocessor, and an audiovisual output assembly. The wearable frame includes rims, a pair of temples, and a pair of lenses in the rims. The audiovisual input assembly is positioned in the frame and is designed to receive spoken and written words. The microprocessor is positioned in the frame and is operatively connected to the audiovisual input assembly to receive input words from the audiovisual input assembly and translate input words into a selected language. The audiovisual output assembly is positioned in the frame and operatively connected to the microprocessor to produce visual text or audio of translated input words received from the microprocessor. The eyeglass translation system can detect spoken or written words, translate the words into a selected language, and then provide audio or visual images of the translated words.

Techniques for performing an action with respect to displayed content are described. A natural language interpretation corresponding to a received spoken user input may be determined. Prior to receiving the spoken user input, content may be displayed to the user from which the spoken user input was received. The natural language interpretation may represent a request to perform an action with respect to a portion of the content currently being displayed. Content identifiers corresponding to content being displayed, may be determined, and embedding data representing at least one feature of the content may be determined using the content identifiers. The natural language interpretation and the embedding data may be processed to determine that the spoken user input relates to a first portion of the displayed content instead of a second portion of the displayed content. Based on the determination, an action responsive to the spoken user input may be performed.

A hybrid, artificial intelligence-based system (100) for integrating speech, web, and hardware for intelligent assistance, wherein the system (100) comprises: a. a speech interaction layer configured to capture a user's speech commands, the speech interaction layer comprising a microphone for receiving spoken input from the user, a speech-to-text engine for converting said spoken input into text, a transformer-based natural language processing module for interpreting an intent of said text, a named entity recognition module for extracting entities including devices, actions, and locations from said text, and a text-to-speech engine for generating spoken responses; b.a web-based visualization layer communicatively coupled to a central command controller, wherein the web-based visualization layer is configured to display live sensor readings, artificial intelligence interpretations, and device states on an interactive dashboard and to maintain synchronization with backend application programming interfaces; c. a hardware control layer communicatively coupled to the central command controller, wherein the hardware control layer comprises one or more microcontroller-based modules configured to manage a variety of sensors and actuators and to communicate with the central command controller via a low-latency, bidirectional communication protocol; d.the central command controller, which is communicatively coupled with the speech interaction layer, the web-based visualization layer, and the hardware control layer, wherein the central command controller is configured to decode the said interpreted intention, generate a structured command packet, and transmit the said structured command packet to the hardware control layer for control and simultaneously to the web-based visualization layer and the speech interaction layer for synchronized multimodal feedback; unde.a secure communication module that is operationally connected to the central command controller and configured to encrypt communication between the said speech interaction layer, the web-based visualization layer, and the hardware control layer, and to authenticate each connected hardware device via a unique hardware identifier; wherein the said synchronized multimodal feedback includes spoken acknowledgment via the text-to-speech engine, visualization of an action and device state on the said interactive dashboard, and execution of a corresponding hardware response by the said hardware control layer.

An input system comprises a spoken input processor configured to receive spoken inputs from a microphone operably coupled to the input system; a physical input processor configured to receive physical inputs from a peripheral device operably coupled to the input system, the physical inputs having a timing relative to the spoken inputs; and a speech recognition processor configured to recognise speech from the speech inputs; wherein the speech recognition processor uses at least some of the received physical inputs, and their timing relative to the spoken inputs, as part of the speech recognition process.

A voice processing system for obtaining data from built-in test equipment of an aircraft which includes an audio device adapted and configured to receive spoken instruction of a user, a voice processor adapted and configured to process the audio input and to translate the audio input into command terms and a memory having a command database programmed with command terms. The memory is adapted and configured to receive translated command terms from audio input, search the command database for translated command terms, map the translated command terms to programmed command terms, and output a command message for retrieval. The system also includes a command processor adapted and configured to receive the command message from the memory and to perform the command by searching stored built-in test equipment failure data in accordance with the command message.

Techniques for performing an action with respect to displayed content are described. A natural language interpretation corresponding to a received spoken user input may be determined. Prior to receiving the spoken user input, content may be displayed to the user from which the spoken user input was received. The natural language interpretation may represent a request to perform an action with respect to a portion of the content currently being displayed. Content identifiers corresponding to content being displayed, may be determined, and embedding data representing at least one feature of the content may be determined using the content identifiers. The natural language interpretation and the embedding data may be processed to determine that the spoken user input relates to a first portion of the displayed content instead of a second portion of the displayed content. Based on the determination, an action responsive to the spoken user input may be performed.