Abstract: There is provided an information presentation device that is configured to present information, to a plurality of users that differ in level, in such a manner that each of the users can easily understand the information, and an information presentation method. The information presentation device includes: an identification unit that identifies respective levels of one or more users; an obtaining unit that obtains presentation information to be presented to the users; a conversion unit that appropriately converts the obtained presentation information according to the level of each user; and a presentation unit that presents the appropriately converted presentation information to each user. The present technology can be applied to, for example, a robot, a signage device, a car navigation device, and the like.

Abstract: Systems and methods are described for disambiguating terms, a challenging problem in computational linguistics. An ambiguous term may be regular words or phrases, or preferably abbreviations or acronyms. A potentially ambiguous term may be identified from an information source, and has two or more potential meanings. For each potential meaning, context and frequency of the potential meaning are determined. Context may include section headings, nearby concepts, or all relevant concepts within the information source. Frequency may reflect frequency of the candidate concepts within literature, medical literature, patient records, or another information source. Context, frequency, or both for each potential meaning can support a computer technology algorithm to select one potential meaning over others.

Abstract: Methods, apparatus, systems, computing devices, computing entities, and/or the like for programmatically generating text strings are provided. An example method may include receiving an input string; generating a first utterance set based at least in part on the input string and a randomness indicator via a generative machine learning model; generating at least one grammar context score for the first utterance set; generating at least one relevance context score for the first utterance set; adjusting the randomness indicator based at least in part on one or more of the at least one grammar context score, a grammar context threshold, the at least one relevance context score, and a relevance context threshold; and generating a second utterance set based at least in part on the input string and the adjusted randomness indicator.

Abstract: The method of the present disclosure includes: obtaining an image to be processed and a question text corresponding to the image; using an optimized dialogue model to encode the image into an image vector and encode the question text into a question vector; generating a state vector based on the image vector and the question vector; decoding the state vector to obtain and output an answer text. A discriminator needs to be introduced in an optimization process of the optimized dialogue model. The dialogue model and the discriminator are alternately optimized until a value of a hybrid loss function of the dialogue model and a value of a loss function of the discriminator do not decrease or fall below a preset value, thereby accomplishing the optimization process.

Abstract: A method includes receiving audio data of an utterance and processing the audio data to obtain, as output from a speech recognition model configured to jointly perform speech decoding and endpointing of utterances: partial speech recognition results for the utterance; and an endpoint indication indicating when the utterance has ended. While processing the audio data, the method also includes detecting, based on the endpoint indication, the end of the utterance. In response to detecting the end of the utterance, the method also includes terminating the processing of any subsequent audio data received after the end of the utterance was detected.

Abstract: A method of transcribing speech using a multilingual end-to-end (E2E) speech recognition model includes receiving audio data for an utterance spoken in a particular native language, obtaining a language vector identifying the particular language, and processing, using the multilingual E2E speech recognition model, the language vector and acoustic features derived from the audio data to generate a transcription for the utterance. The multilingual E2E speech recognition model includes a plurality of language-specific adaptor modules that include one or more adaptor modules specific to the particular native language and one or more other adaptor modules specific to at least one other native language different than the particular native language. The method also includes providing the transcription for output.

Abstract: A quantization apparatus comprises: a first quantization module for performing quantization without an inter-frame prediction; and a second quantization module for performing quantization with an inter-frame prediction, and the first quantization module comprises: a first quantization part for quantizing an input signal; and a third quantization part for quantizing a first quantization error signal, and the second quantization module comprises: a second quantization part for quantizing a prediction error; and a fourth quantization part for quantizing a second quantization error signal, and the first quantization part and the second quantization part comprise a trellis structured vector quantizer.

Abstract: A method for detecting a voice activity in an input audio signal composed of frames includes that a noise characteristic of the input signal is determined based on a received frame of the input audio signal. A voice activity detection (VAD) parameter is derived based on the noise characteristic of the input audio signal using an adaptive function. The derived VAD parameter is compared with a threshold value to provide a voice activity detection decision. The input audio signal is processed according to the voice activity detection decision.

Abstract: Knowledge gaps in a chatbot are identified with reference to a domain-specific document and a set of QA pairs of the chatbot. Entities and/or entity values associated with the document are compared to the entities and/or entity values of the QA pairs. Entities of the document not associated with the QA pairs are identified as knowledge gaps. The QA pairs and knowledge gaps are ranked by relevance to the domain.

Abstract: An audio processing unit (APU) is disclosed. The APU includes a buffer memory configured to store at least one frame of an encoded audio bitstream, where the encoded audio bitstream includes audio data and a metadata container. The metadata container includes a header and one or more metadata payloads after the header. The one or more metadata payloads include dynamic range compression (DRC) metadata, and the DRC metadata is or includes profile metadata indicative of whether the DRC metadata includes dynamic range compression (DRC) control values for use in performing dynamic range compression in accordance with at least one compression profile on audio content indicated by at least one block of the audio data.

Abstract: Aspects of the disclosure relate to systems and methods for increasing the speed, accuracy, and efficiency of language processing systems. A provided method may include storing a plurality of modules in a database. The method may include configuring the plurality of modules in a multi-tier tree architecture. The method may include receiving an utterance. The method may include processing the utterance via a natural language processing (NLP) engine. The method may include routing the utterance. The routing may include identifying a highest tier module that matches a predetermined portion of the utterance. The method may include compiling a result set of modules. The method may include transmitting the result set of modules to the system user. The result set of modules may include a comprehensive and narrowly tailored response to the user request.

Abstract: An exemplary speech presentation system receives a simulated binaural audio signal merging together a plurality of concurrent speech instances originating from a plurality of different speakers speaking concurrently. The speech presentation system also receives acoustic propagation data representative of respective propagation effects applied, within the simulated binaural audio signal, to each of the concurrent speech instances to simulate propagation of the concurrent speech instances. Based on the acoustic propagation data, the speech presentation system extracts, from the simulated binaural audio signal, a different auto-transcribable speech signal for each of the plurality of concurrent speech instances. Additionally, based on the extracted auto-transcribable speech signals, the speech presentation system generates a different closed captioning dataset for each of the plurality of concurrent speech instances. Corresponding methods and systems are also disclosed.

Abstract: An output method includes obtaining voice information, determining whether the voice information is a voice request, in response to the voice information being the voice request, obtaining reply information for replying to the voice request, and supplemental information, and transmitting the reply information and the supplementary information to an output device for outputting. The supplemental information is information that needs to be outputted in association with the reply information.

Abstract: Systems and methods are disclosed for end-to-end neural networks for speech recognition and classification and additional machine learning techniques that may be used in conjunction or separately. Some embodiments comprise multiple neural networks, directly connected to each other to form an end-to-end neural network. One embodiment comprises a convolutional network, a first fully-connected network, a recurrent network, a second fully-connected network, and an output network. Some embodiments are related to generating speech transcriptions, and some embodiments relate to classifying speech into a number of classifications.

Abstract: Techniques for routing a user command to a speechlet and resolving conflicts between potential speechlets are described. A system determines an intent of an input command. The system also receives context information associated with the input command. The system determines speechlets (e.g., speechlets and/or skills) that may execute with respect to the input command given the intent and the context data. The system then determines whether conditions of routing rules, associated with the speechlets, are satisfied given the context data. If the conditions of only one routing rule are satisfied, the system causes the speechlet associated with the routing rule to execute with respect to the input command. If the conditions of more than one routing rule are satisfied, the system may determine a speechlet to execute with respect to the input command based on the speechlets' priorities in a list of speechlets and/or based on potential output data provided by the speechlets.

Abstract: Techniques described herein relate to facilitating end-to-end multilingual communications with automated assistants. In various implementations, speech recognition output may be generated based on voice input in a first language. A first language intent may be identified based on the speech recognition output and fulfilled in order to generate a first natural language output candidate in the first language. At least part of the speech recognition output may be translated to a second language to generate an at least partial translation, which may then be used to identify a second language intent that is fulfilled to generate a second natural language output candidate in the second language. Scores may be determined for the first and second natural language output candidates, and based on the scores, a natural language output may be selected for presentation.

Abstract: Systems and methods for automatically selecting a smart device based on the context of a voice command event are disclosed. In embodiments, a smart device controller receives a user voice command from a user regarding a function to be performed during a command event within a smart environment; determines context data associated with the user voice command; determines a smart device name of the user voice command; accesses a knowledge database including registered smart devices and associated context-specific names, wherein the context-specific names are associated with context parameters; matches the smart device name of the user voice command with one of the context-specific names of the knowledge database; automatically chooses a select smart device from the plurality of registered smart devices to perform the function based on the matching; and initiates the function at the select smart device.

Abstract: A method for summarizing text with sentence extraction including steps as follows. Sentences are extracted from a document including text by a natural language processing (NLP) based feature extractor. A word vector set with respect to each of the sentences is generated by a processor. The word vector set with respect to each of the sentences is used to generate a n-grams vector set and a phrase-n vector set with respect to each of the sentences. A word score representing similarity between the word vector sets, a n-grams score representing similarity between the n-grams vector sets, and a phrase-n score representing similarity between the phrase-n vector sets are computed. The word, n-grams, and phrase-n scores are combined to compute an edge score. Text features are selected from the sentences using the edge scores of the sentences, so as to output a summary of the document.

Abstract: An artificial intelligence (AI) device may acquire a probability that a received speech signal is classified as a noise signal, calculate a confidence level of a first model for determining to which phoneme the speech signal belongs, based on the speech signal, determine a weight of the first model based on the probability and the confidence level of the first model, and output a speech recognition result of the speech signal using the determined weight of the first model.

Abstract: An audio packet error concealment system includes an encoding unit for encoding an audio signal consisting of a plurality of frames, and an auxiliary information encoding unit for estimating and encoding auxiliary information about a temporal change of power of the audio signal. The auxiliary information is used in packet loss concealment in decoding of the audio signal. The auxiliary information about the temporal change of power may contain a parameter that functionally approximates a plurality of powers of subframes shorter than one frame, or may contain information about a vector obtained by vector quantization of a plurality of powers of subframes shorter than one frame.