Abstract: Embodiments are provided for suggesting topics in a messaging system. A set of queries is received from a chat transcript history, where the set of queries includes a set of unhandled queries, and each unhandled query comprises a query for which a bot did not identify a corresponding topic (e.g., queries that did not trigger selection of a topic by the bot). A vector representation is generated for each unhandled query in the set of unhandled queries. The vector representations for the set of unhandled queries are clustered to generate one or more clusters of vector representations, each cluster corresponding to a group of unhandled queries. A corresponding suggested topic is generated for each cluster and provided to an authoring tool that comprises one or more interactive elements to enable an author to select at least one of the suggested topics for implementation in the bot.

Abstract: Embodiments are disclosed for bitrate distribution in immersive voice and audio services. In an embodiment, a method of encoding an IVAS bitstream comprises: receiving an input audio signal; downmixing the input audio signal into one or more downmix channels and spatial metadata; reading a set of one or more bitrates for the downmix channels and a set of quantization levels for the spatial metadata from a bitrate distribution control table; determining a combination of the one or more bitrates for the downmix channels; determining a metadata quantization level from the set of metadata quantization levels using a bitrate distribution process; quantizing and coding the spatial metadata using the metadata quantization level; generating, using the combination of one or more bitrates, a downmix bitstream for the one or more downmix channels; combining the downmix bitstream, the quantized and coded spatial metadata and the set of quantization levels into the IVAS bitstream.

Abstract: A method of training a ranking model, and an electronic device, which relate to technical fields of natural language processing and intelligent search. The method includes: in training the ranking model, firstly acquiring a plurality of first sample pairs and respective label information; for each first sample pair, inputting a first search text, a first title text of a first candidate text, and a first target summary corresponding to the first candidate text into an initial language model to obtain a second relevance score corresponding to the each first sample pair; then using the first target summary to replace the first candidate text to participate in the training of the ranking model, and updating at least one network parameter of the initial language model according to the label information and the second relevance score corresponding to each first sample pair.

Abstract: Systems and methods are provided for obtaining, training, and using an end-to-end AST model based on a neural transducer, the end-to-end AST model comprising at least (i) an acoustic encoder which is configured to receive and encode audio data, (ii) a prediction network which is integrated in a parallel model architecture with the acoustic encoder in the end-to-end AST model, and (iii) a joint layer which is integrated in series with the acoustic encoder and prediction network. The end-to-end AST model is configured to generate a transcription in the second language of input audio data in the first language such that the acoustic encoder learns a plurality of temporal processing paths.

Abstract: In an audio signal, one or more processing circuits recognize spoken content in a user's own speech signal using speech recognition and natural language understanding. The spoken content describes a listening difficulty of the user. The one or more processing circuits generate, based on the spoken content, one or more actions for hearing devices and feedback for the user. The one or more actions attempt to resolve the listening difficulty. Additionally, the one or more processing circuits convert the user feedback to verbal feedback using speech synthesis and transmit the one or more actions and the verbal feedback to the hearing devices via a body-worn device. The hearing devices are configured to perform the one or more actions and play back the verbal feedback to the user.

Abstract: A text reconstruction system for reconstructing a primary text data is provided. A voice input signal of a user is converted into an input text data by a speech recognition module. A text classifier module generates one or more tokens and adds the tokens into a word bag corresponding to the user. A text identifier module generates a text corpus based on the input text data. A user profile builder module creates a user profile based on the word bag, the input text data, and the text corpus. A decision module determines, based on the word bag, whether the primary data is to be reconstructed and reconstructs the primary text data to generate a personalized text data based on the user profile.

Abstract: A method for interpreting structured and unstructured content to facilitate tailored transactions is provided. The method includes acquiring, from one or more unstructured data sources, unstructured data, and acquiring, from one or more structured data sources, structured data. The method further includes performing natural language processing (NLP) on both the structured data and the unstructured data using a machine learning algorithm, and generating, via the machine learning algorithm, an NLP response based on the NLP. Based on the NLP response, the method further performs identifying at least one candidate object, and generating a list of actions corresponding to the candidate object.

Abstract: One example method includes performing sound quality operations. Microphone arrays are used to cancel background noise and to enhance speech. With arrays at each environment of each user participating in a call, a first microphone array can cancel or suppress background noise and a second array can generate enhanced speech for transmission to other users. Thus, for user, the audio signal output by the user's device includes an anti-noise signal to cancel background noise present in the user's environment and enhanced speech from other users.

Abstract: Disclosed are a method for correcting a text, an electronic device and a storage medium. The method includes: acquiring a text to be corrected; acquiring a phonetic symbol sequence of the text to be corrected; and obtaining a corrected text by inputting the text to be corrected and the phonetic symbol sequence into a text correction model, in which, the text correction model obtains the corrected text by: detecting an error word in the text to be corrected, determining a phonetic symbol corresponding to the error word in the phonetic symbol sequence, and adding the phonetic feature corresponding to the phonetic symbol behind the error word to obtain a phonetic symbol text, and correcting the error word and the phonetic feature in the phonetic symbol text to obtain the corrected text.

Abstract: Systems and processes for speech interpretation based on environmental context are provided. For example, a user gaze direction is detected, and a speech input is received from a first user of the electronic device. In accordance with a determination that the user gaze is directed at a digital assistant object, the speech input is processed by the digital assistant. In accordance with a determination that the user gaze is not directed at a digital assistant object, contextual information associated with the electronic device is obtained, wherein the contextual information includes speech from a second user. Determination is made whether the speech input is directed to a digital assistant of the electronic device. In accordance with a determination that the speech input is directed to a digital assistant of the electronic device, the speech input is processed by the digital assistant.

Abstract: A method of noise reduction, which is applied to an electronic device. The electronic device includes a first sound collector and a second sound collector, installation positions of the first sound collector and the second sound collectors are different; the method includes: determining a desired sound signal and an interference sound signal based on a first sound signal collected by the first sound collector and a second sound signal collected by the second sound collector (S102); obtaining a third sound signal by performing coherent noise elimination processing on the desired sound signal based on the interfering sound signal (S103); and then obtaining a target sound signal by performing incoherent noise suppression processing on the third sound signal based on a probability of existence of a speech in the third sound signal (S104).

Abstract: A system and method for temporarily disabling keyword detection to avoid detection of machine-generated keywords. A local device may operate two keyword detectors. The first keyword detector operates on input audio data received by a microphone to capture keywords uttered by a user. In these instances, the keyword may be detected by the first detector and the audio data may be indicated for speech processing. The system may determine output audio data responsive to the input audio data. The local device may process the output audio data to determine that it also includes the keyword. The device may then disable the first keyword detector while the output audio data is played back by an audio speaker of the local device. Thus the local device may avoid detection of a keyword originating from the output audio. The first keyword detector may be reactivated after a time interval during which the keyword might be detectable in the output audio.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for speech recognition using neural networks. A feature vector that models audio characteristics of a portion of an utterance is received. Data indicative of latent variables of multivariate factor analysis is received. The feature vector and the data indicative of the latent variables is provided as input to a neural network. A candidate transcription for the utterance is determined based on at least an output of the neural network.

Abstract: A system and method for real-time fraud detection with a social engineering phoneme (SEP) watchlist of phoneme sequences may perform real-time fraud prevention operations including receiving incoming call interactions and grouping the call interactions into one or more clusters, each cluster associated with a speaker's voice based on voiceprints. For a pair of voiceprints in a cluster, a phoneme sequence is extracted for each voice print. From the extracted phoneme sequences, a similarity score is then calculated to determine if a match exists between the extracted phoneme sequences based on a threshold. If determined a match exists, the phoneme sequence may be added to a SEP watchlist.

Abstract: One embodiment provides a method, including: receiving, at an information handling device, text associated with a user command; storing, in a data store, an encrypted form of the text associated with the user command; determining, using a processor, whether the encrypted form of the text has been detected in other user commands in exceedance of a predetermined threshold; and storing, responsive to determining that the encrypted form of the text has been detected in the other user commands in exceedance of the predetermined threshold, an unencrypted transcript of the text in a data table. Other aspects are described and claimed.

Abstract: An incident report management system is configured to receive and analyze incident reports relating to workplace accidents and injuries. A natural language processing function utilizes word dictionaries of varying type and scope to reduce an incident description to a set of core components that are far smaller than the input text while also preserving important aspects of the input text. The reduced core component set may be analyzed for meaning, compared to large volumes of historic incident reports, and otherwise processed more quickly and more efficiently whether by an expert function or AI function. In this manner, the system is able to provide real-time feedback during submission of incidents to improve quality and completeness, and after submission of incidents to notify users of serious incidents, provide dashboard analytics, and identify underlying and undiscovered risks in the workplace.

Abstract: The technology relates to systems and methods for transcribing audio of a meeting. Upon transcribing the audio, the systems and methods can parse different portions of the prescribed audio so that they may attribute the different portions to a particular speaker. These transcribed portions that are attributed to a particular speaker are made available for viewing and interacting using a graphical user interface.

Abstract: Example methods and systems are directed to automatic data card generation for datasets. A data card is a summary that describes quantitative aspects of a dataset, qualitative aspects of a dataset, or both. The data samples and documentation of a dataset are analyzed automatically to determine a number of samples, a primary data type, a license, or any suitable combination thereof. Data formats for data and documentation of the dataset may be automatically recognized. Language of text data may be automatically recognized. The most frequent language for the text data may be identified as the primary language of the dataset. A data card may be created for the dataset. The data card may indicate the number of samples, the data formats used in the data set, the language of text data in the dataset, or any suitable combination thereof.

Abstract: A method for constructing a decoding network, a speech recognition method, a device, an apparatus, and a storage medium are provided. The method for constructing a decoding network includes: acquiring a general language model, a domain language model, and a general decoding network generated based on the general language model; generating a domain decoding network based on the domain language model and the general language model; and integrating the domain decoding network with the general decoding network to obtain a target decoding network. The speech recognition method includes: decoding to-be-recognized speech data by using a target decoding network to obtain a decoding path for the to-be-recognized speech data; and determining a speech recognition result for the to-be-recognized speech data based on the decoding path for the to-be-recognized speech data.

Abstract: Described herein is a method of encoding an audio signal. The method comprises: generating a plurality of subband audio signals based on the audio signal; determining a spectral envelope of the audio signal; for each subband audio signal, determining autocorrelation information for the subband audio signal based on an autocorrelation function of the subband audio signal; and generating an encoded representation of the audio signal, the encoded representation comprising a representation of the spectral envelope of the audio signal and a representation of the autocorrelation information for the plurality of subband audio signals. Further described are methods of decoding the audio signal from the encoded representation, as well as corresponding encoders, decoders, computer programs, and computer-readable recording media.