Abstract: A real-time processor-implemented translation method and apparatus is provided. The real-time translation method includes receiving a content, determining a delay time for real-time translation based on a silence interval of the received content and an utterance interval of the received content, generating a translation result by translating a language used in the received content, and synthesizing the translation result and the received content.

Abstract: A speech recognition method includes receiving speech data, obtaining, from the received speech data, a candidate text including at least one word and a phonetic symbol sequence associated with a pronunciation of a target word included in the received speech data, using a speech recognition model, replacing the phonetic symbol sequence included in the candidate text with a replacement word corresponding to the phonetic symbol sequence, and determining a target text corresponding to the received speech data based on a result of the replacing.

Abstract: Systems, devices and methods for delivering audible alerts are described. In one aspect, an event for generating an audible alert for a user is detected. In response to detection of the event for generating the audible alert, it is determined whether an authorized audible interface device is within a threshold distance of the user based on a location of the user and a location of one or more authorized audible interface devices. In response to determining that an audible interface device is within the threshold distance of the user, alert instructions for the audible interface device are generated. The alert instructions cause the audible interface device to generate the audible alert in accordance with alert data provided in the alert instructions. The alert instructions are sent to the audible interface device over a wireless network via a communication module.

Abstract: A security panel for controlling home automation devices via a voice assistant device is provided, in which the security panel includes a processor, a memory, a microphone, and a speaker. In one example implementation, the security panel is configured to receive a text input from a user, convert the text input into an audio format via a text-to-speech engine to generate a first voice command for controlling one or more home automation devices via a voice assistant device, and to output the first voice command via the speaker of the security panel, in which the first voice command is received by the voice assistant device via a microphone of the voice assistant device, in which the voice assistant device is configured to control the one or more home automation devices based on the first voice command.

Abstract: Embodiments of the present disclosure include systems and methods for packing tokens to train sequence models. In some embodiments, a plurality of datasets for training a sequence model is received. Each dataset in the plurality of datasets includes a sequence of correlated tokens. A set of training data is generated that includes a subset of a sequence of tokens from a first dataset in the plurality of datasets and a subset of a sequence of tokens from a second, different dataset in the plurality of datasets. The sequence model is trained using the set of training data.

Abstract: A method, system, and non-transitory computer readable media for converting input from a user into a human interface device (HID) output to cause a corresponding action at an mapped device includes receiving one or more user input from a user at an input device, analyzing the user input selecting a command from a command profile that maps at least one of the received user inputs to one or more mapped tasks, executing the one or more mapped tasks associated with the selected command, and causing one or more corresponding actions at one or more mapped devices associated with the one or more mapped tasks.

Abstract: Various embodiments described herein relate to determining and providing user-specific feedback based on an analysis of audible input sessions performed by a user. In this regard, a set of term recognition structures that each comprise a plurality of term data objects and a respective confidence score for each term data object are generated. For at least one pairing of term data objects of a predefined term glossary, a correlation coefficient value for the respective pairing is determined. In accordance with determining that the correlation coefficient value for the at least one pairing satisfies a predefined threshold a generate a visualization is generated and displayed that includes an indication of the term data objects of the at least one pairing.

Abstract: A computer implemented method, computer system, and computer program product for executing a voice command. A number of processor units displays a view of a location with voice command devices in response to detecting the voice command from a user. The number of processor units displays a voice command direction for the voice command in the view of the location. The number of processor units changes the voice command direction in response to a user input. The number of processor units identifies a voice command device from the voice command devices in the location based on the voice command direction to form a selected voice command device. The number of processor units executes the voice command using the selected voice command device.

Abstract: Methods, systems, computer program products and data structures are described which allow for efficient correction of a transcription output of an automatic speech recognition system by a human proofreader. A method comprises receiving a voice input from a user; determining a transcription of the voice input; providing the transcription of the voice input; receiving a text input from the user indicating a revision to the transcription; determining how to revise the transcription in accordance with the text input; and revising the transcription of the voice input in accordance with the text input. A general or specialized language model, an acoustical language model, a character language model, a gaze tracker, and/or a stylus may be used to determine how to revise the transcription in accordance with the text input.

Abstract: Implementations set forth herein relate to an automated assistant capable of bypassing soliciting a user for supplemental data for completing an action when a previously-queried application is capable of providing the supplemental data. For instance, when a user invokes the automated assistant to complete a first action with a first application, the user may provide many pertinent details. Those details may be useful to a second application that the user may subsequently invoke via the automated assistant for completing a second action. In order to save the user from having to repeat the details to the automated assistant, the automated assistant can interact with the first application in order to obtain any information that may be essential for the second application to complete the second action. The automated assistant can then provide the information to the second application, without soliciting the user for the information.

Abstract: The method system described herein is configured to identify and execute a next recommended action for a user based on the user's audio input. In an embodiment, the system is configured to receive audio input from the user and convert the audio input into a string. The system may identify an attribute associated with the user. The system may identify a type of action based on the string. The system may query a data repository using the attribute associated with the user to retrieve information associated with the type of action. The system may identify a recommended action for the user based on the information associated with the type of action. The system may then execute the recommended action for the user.

Abstract: A learning device calculates an image feature using a model (image encoder) that receives an image and outputs the image feature obtained by mapping the image into a latent space. The learning device calculates an audio feature using a model (audio encoder) that receives a speech in a predetermined language and outputs the audio feature obtained by mapping the speech into the latent space, and that includes a neural network provided with a self-attention mechanism. The learning device updates parameters of the models used by an image feature calculation unit and an audio feature calculation unit such that the image feature of a first image is similar to the audio feature of a speech corresponding to the first image.

Abstract: A phonetic search system may pass phonetic information from an automatic speech recognition (ASR) system to a natural language understanding (NLU) system for the latter to leverage when performing entity resolution in the presence of ambiguous interpretations. The ASR system may include an acoustic model and a language model. The acoustic model can process audio data to generate hypotheses that can be mapped to acoustic data; i.e., one or more acoustic units such as phonemes. The language model can process the acoustic units to generate text data representing possible transcriptions of the audio data. ASR/NLU systems may have difficulty interpreting speech when confronted with, for example, homographs, which are words that are spelled the same, but have different meanings. When uncertainty in the final transcription is high, the system can leverage the acoustic data to improve the accuracy of entity resolution.

Abstract: An example process includes: receiving a first natural language input; initiating, by a digital assistant operating on the electronic device, a first task based on the first natural language input; determining whether the first task is of a predetermined type; and in accordance with a determination that the first task is of a predetermined type: determining whether one or more criteria are satisfied; and providing a response to the first natural language input, where providing the response includes: in accordance with a determination that the one or more criteria are not satisfied, outputting a first sound indicative of the initiated first task and a first verbal response indicative of the initiated first task; and in accordance with a determination that the one or more criteria are satisfied, outputting the first sound without outputting the first verbal response.

Abstract: A method for enhancing telephone speech signals based on Deep Convolutional Neural Network (CNN) is disclosed. The method is able to reduce the effect of acoustic distortions in daily scenarios during a telephone call. It is a single-channel, speech-oriented method with causal design and low latency. The novelty lies in the noise reduction method which, based on the classical gain method, uses a CNN to learn the Wiener estimator. Then, it computes the gain of the filter to enhance the speech power over the noise power for each time-frequency component of the signal. The selection of the Wiener gain estimator as an essential element of the method, decreases the vulnerability to estimation errors since the characteristics of this measure make it very appropriate to be estimated by deep learning approaches.

Abstract: An agent device that receives, from an onboard device installed in a vehicle, vehicle information relating to the vehicle and question information corresponding to a question from a user, based on the vehicle information, confirms a scope of questions for which generation of a response is not possible, and instructs the onboard device to block receipt of questions falling within the scope of questions for which it has been confirmed that response generation is not possible.

Abstract: A voice command can be received from a user. One or more voice command devices (VCDs) that the voice command is targeting can be determined. A visual indicator of each of the one or more targeted VCDs can be displayed on an XR device worn by the user, wherein each visual indicator visually indicates a respective targeted VCD the voice command is directed to on the XR device.

Abstract: Electronic device includes display, microphone, and processor configured to activate voice input function based on user input, display graphic representation for indicating that the voice input function is activated, provide, on the display, a text display area for displaying text inputted by a plurality of user input methods and a keyboard input interface for receiving a user keyboard input, the plurality of user input methods including user keyboard input method and user voice input method, receive, via the keyboard input interface, the user keyboard input corresponding to a first text, display the first text in the text display area based on receiving the user keyboard input, receive user voice input corresponding to a second text while the keyboard input interface is provided and the voice input function is activated, and display the second text next to the first text in the text display area based on the user voice input.

Abstract: A method is provided for generating a ranked list of candidate responders. In some embodiments, the method includes receiving a question from a user and generating a question feature vector representing an intent of the question and a first skill set inferred from the question. The method also includes for one or more candidate responders, generating a candidate feature vector representing a skill set and questions associated with the respective candidate responder; computing a reputation score based on questions and user feedback associated with the respective candidate responder; and computing, based on the question feature vector, candidate feature vector, and reputation score, a probability score representing a prediction of the quality of an answer that would be provided by the respective candidate responder if the input question were routed to the respective candidate responder. The method further includes generating a ranked list of candidate responders using the computed probability scores.

Abstract: An audio signal processing method receives, by a terminal, a backtalk input instruction from a performer, obtains, by a microphone connected to the terminal, voice information from the performer, and outputs, in a case where the backtalk input instruction has been received by the terminal, a backtalk signal corresponding to the voice information obtained by the microphone connected to the terminal to a monitor bus of a mixer.