Abstract: Systems and methods to analyze and identify adaptations to document based on an audience of the document to improve a document quality are disclosed. Exemplary implementations may: analyze individual documents to determine its active lexicon from lexicons stored in electronic storage that the individual documents are comprised of; store the documents in correlation with the determined active lexicons in the electronic storage; generate, based on the lexicons and the correlations, audience scores for individual ones of the documents; store the audience scores in the electronic storage; receive a request to obtain a document; and effectuate presentation of the document and the audiences scores corresponding to the document.

Abstract: Systems and methods for determining, based on invocation input that is common to multiple automated assistants, which automated assistant to invoke in lieu of invoking other automated assistants. The invocation input is processed to determine one or more invocation features that may be utilized to determine which, of a plurality of candidate automated assistants, to invoke. Further, additional features are processed that can indicate which, of the plurality of invocable automated assistants, to invoke. Once an automated assistant has been invoked, additional audio data and/or features of additional audio data are provided to the invoked automated assistant for further processing.

Abstract: The present disclosure relates to an audio signal processing method implemented by an audio system which includes at least an internal sensor, wherein the internal sensor is located in an ear canal of a user of the audio system and arranged to measure acoustic signals which propagate internally to a head of the user. The audio signal processing method includes: producing an internal audio signal by the internal sensor; determining an audio spectrum of the internal audio signal; determining a spectral center of the audio spectrum; determining a spectrum shape correction filter based on the spectral center; and filtering the internal audio signal by using the spectrum shape correction filter, thereby producing a filtered internal audio signal.

Abstract: A method for generating a first case dataset in a first language. The method includes receiving adverse event data. The method further includes determining case data including general case data and regional case data and providing the case data to a translator computing device to enable display on a user interface including multiple duolingual text fields with a first language text field including at least a portion of the text data in the first language and a second language text field adjacent the first language text field. The method further includes receiving the text data in the second language from a translator computing device. The text data in the second language is received via the second language text fields of the plurality of duolingual text fields. The method further includes generating and outputting the first case dataset including the text data in the first language.

Abstract: An utterance analysis device including; a calculator that receives utterance data in order of time series and analyzes a change in the utterances by using a plurality of first likelihoods, which are each values for identifying a possibility that the utterance data corresponds to each category; and a display processor that displays visualization data visualizing the change in the utterances which is obtained by the calculator. The calculator determines a second likelihood for each category by integrating the first likelihoods of a plurality of pieces of utterance data in a second period which is shorter than the first period. The visualization data displayed by the display processor represents a change in the utterances based on a change in the second likelihood of each category in a plurality of different second periods.

Abstract: A system and method for analyzing a textual narrative are provided. The method includes: receiving, from a user, an input that includes a textual narrative; comparing the textual narrative to various types of quality criteria, including syntactical criteria, semantic criteria, and pragmatic criteria; determining, based on the comparison, whether the textual narrative satisfies the quality criteria; and providing an output that indicates a result of the determination. The textual narrative may be a JIRA user story or a JIRA epic. The comparison may be performed by using one or more Natural Language Processing techniques.

Abstract: Generating text suggestions based on context can leverage sources associated with the context to generate more accurate and informed text suggestions. For example, the context can be a user situation, such as the user is attending a meeting. Obtaining text from sources associated with the user situation can generate a corpus of text that can be leveraged for generating the context-based text suggestions.

Abstract: A method and system are provided to facilitate statement acquisition and verification pertaining to an incident. One or more processors are operationally coupled to one or more software modules, configured to: receive audio data between a suspect and an officer at a portable radio, detect a pre-Miranda statement made by the suspect within the audio data; detect a Miranda warning made by the officer within the audio data; detect a waiving of Miranda rights by the suspect within the audio data; detect a post-Miranda statement by the suspect within the audio data after the waiving of Miranda rights; determine when a degree of semantic similarity between the pre-Miranda statement and the post-Miranda statement fails to satisfy an admissibility condition; generate one or more questions to induce the suspect to repeat the pre-Miranda statement; and present the one or more questions to an electronic user interface.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for using a generative neural network to convert conditioning text inputs to audio outputs. The generative neural network includes an alignment neural network that is configured to receive a generative input that includes the conditioning text input and to process the generative input to generate an aligned conditioning sequence that comprises a respective feature representation at each of a plurality of first time steps and that is temporally aligned with the audio output.

Abstract: A method of training a model, a method of determining a word vector, a device, a medium, and a product are provided, which may be applied to fields of natural language processing, information processing, etc. The method includes: acquiring a first word vector set corresponding to a first word set; and generating a reduced-dimensional word vector for each word vector in the first word vector set based on a word embedding model, generating, for other word vector in the first word vector set, a first probability distribution in the first word vector set based on the reduced-dimensional word vector, and adjusting a parameter of the word embedding model so as to minimize a difference between the first probability distribution and a second probability distribution for the other word vector determined by a number of word vector in the first word vector set.

Abstract: Aspects of the present disclosure relate to a well analog recommendation system. The well analog recommendation system may generate numerical representations indicative of text-based descriptions within a well report and/or well log associated with a well. Further, the well analog recommendation system may generate a well analog output that may include one or more text-based characterizations associated with one or more additional wells that are determined based on the numerical representation. For example, the well analog recommendation system may compare the numerical representation of the well to one or more numerical representations associated with the one or more additional wells and output the one or more text-based characterizations when the numerical representations are approximately equal or above a threshold.

Abstract: A de-coupled computing infrastructure is described that is adapted to provide domain specific contextual engines based on conversational flow. The computing infrastructure further includes, in some embodiments, a mechanism for directing conversational flow in respect of a backend natural language processing engine. The computing infrastructure is adapted to control or manage conversational flows using a plurality of natural language processing agents.

Abstract: A system and method for generating task-specific text by processing multimodal inputs using machine-learning models is provided. The method may include accessing first sets of tokens associated with a desired task and one or more modalities associated with a context of the desired task. The method may further include determining a second set of tokens for each of the one or more modalities using a classifier network associated with the modality. The method may further include generating a number of embedding vectors by mapping the first sets of tokens and the second set of tokens associated with each of the one or more modalities to an embedding space. The method may further include producing a sequence of words addressing the desired task by processing the number of embedding vectors with an encoder-decoder network.

Abstract: A method for training a language model comprises (a) receiving vectorized training data as input to a multitask pretraining problem; (b) generating modified vectorized training data based on the vectorized training data, according to an upstream data embedding; (c) emitting pretraining output based on the modified vectorized training data, according to a downstream data embedding equivalent to the upstream data embedding; and (d) adjusting the upstream data embedding and the downstream data embedding by computing, based on the pretraining output, a gradient of the upstream data embedding disentangled from a gradient of the downstream data embedding, thereby advancing the multitask pretraining problem toward a pretrained state.

Abstract: A speech recognition method, including acquiring first linear frequency spectrums corresponding to audios to be trained with different sampling rates; determining the maximum sampling rate and other sampling rates; determining the maximum frequency domain sequence number of the first linear frequency spectrums as a first frequency domain sequence number and a second frequency domain sequence number; in the first linear frequency spectrums corresponding to the other sampling rate, configuring amplitude values corresponding to each frequency domain sequence number that is greater than the first frequency domain sequence number and less than or equal to the second frequency domain sequence number to be zero to obtain second linear frequency spectrums; determining first speech features and second voice features; and using the first speech features and the second speech features to train a machine learning model.

Abstract: An example apparatus for reducing to reduce noise in audio includes a preprocessor to receive audio input from a microphone and preprocess the audio input to generate preprocessed audio. The apparatus also includes an acoustic event detector to detect an acoustic event corresponding to a disturbance in the preprocessed audio. The apparatus further includes a noise reduction model selector to select a specific disturbance model based on the detected acoustic event. The apparatus further includes a noise suppressor to attenuate components related to the disturbance in the preprocessed audio using the selected specific disturbance model to generate enhanced audio with suppressed noise.

Abstract: The present disclosure provides a translation method and apparatus, an electronic device, and a non-transitory storage medium. An implementation includes: determining an encoded feature of a sentence to be translated by an encoding module; determining, by a graph network module, a knowledge fusion feature of the sentence to be translated based on a preset graph network, wherein the preset graph network is constructed based on a polysemous word in a source language corresponding to the sentence to be translated and a plurality of translated words corresponding to the polysemous word in a target language; determining, by a decoding network, a translated sentence corresponding to the sentence to be translated based on the encoded feature and the knowledge fusion feature.

Abstract: Implementations disclosed herein are directed to federated learning of machine learning (“ML”) model(s) based on gradient(s) generated at corresponding client devices and a remote system. Processor(s) of the corresponding client devices can process client data generated locally at the corresponding client devices using corresponding on-device ML model(s) to generate corresponding predicted outputs, generate corresponding client gradients based on the corresponding predicted outputs, and transmit the corresponding client gradients to the remote system. Processor(s) of the remote system can process remote data obtained from remote database(s) using global ML model(s) to generate additional corresponding predicted outputs, generate corresponding remote gradients based on the additional corresponding predicted outputs. Further, the remote system can utilize the corresponding client gradients and the corresponding remote gradients to update the global ML model(s) or weights thereof.

Abstract: Disclosed embodiments may include a method that includes receiving description data in an originating language for a user and a location associated with the user, identifying one or more names from the description data, retrieving additional data in the originating language based on the one or more names, generating enhanced description data in the originating language for the user based on the description data and the additional data, identifying a target language based on the location associated with the user, selecting a first trained neural network from a plurality of trained neural networks based the target language, providing the enhanced description data in the originating language to the first trained neural network, translating, via the first trained neural network, the enhanced description data from the originating language to the target language, and generating a graphical user interface for display that comprises the enhanced description data in the target language.

Abstract: Methods and systems are disclosed configured to perform automated volume leveling on speech content in an audio file containing speech and non-speech segments. A low pass filter and a high pass filter may be applied to the audio data, and normalization may be performed. Speech and non-speech segments may be detected. Gain adjustments may be made to achieve a substantially constant short term loudness. Processing may be applied to enhance speech parameters, such as attack and release. An upward expander may be used to achieve a target loudness level. A limited and/or dynamic range compressor may be utilized to satisfy true peak and/or short term loudness specifications. A file of processed audio data may be generated and transmitted to one or more destinations for broadcast and/or streaming.