Abstract: Image-based machine learning approaches are used to classify audio data, such as speech data as authentic or otherwise. For example, audio data can be obtained and a visual representation of the audio data can be generated. The visual representation can include, for example, an image such as a spectrogram or other visual or electronic representation of the audio data. Before processing the image, the audio data and/or image may undergo various preprocessing techniques. Thereafter, the image representation of the audio data can be analyzed using a trained model to classify the audio data as authentic or otherwise.

Abstract: Methods for calculating nutrient content information. In one embodiment, the methods comprise: receiving a recipe having a list of ingredients and quantities, for each of the ingredients a corresponding record is found within a database of known records, the records are associated to quantities and nutritional values. The units of measurement of the recipe ingredients and the identified record are compared. When the units are the same, no conversion is performed. When the units are different, the units of the known record are converted using a conversion factor derived from a relationship between the differing units of measurement. In one variant, the conversion factor may be identified from a table of conversion factors relating various units of measurement to one another. Finally, the converted or the known nutritional values are multiplied by a ratio of the quantity of the ingredient in the recipe to the quantity of the known record.

Abstract: A method and system for improving the quality of audio communications as perceived by humans include audio signal spectrum frequency shift for enhancement of speech recognition by human customers, including mitigation of common age-related hearing loss on high audio frequencies.

Abstract: The technology described in this document can be embodied in a computer-implemented method that includes receiving, at a processing system, a first signal including an output of a speaker device and an additional audio signal. The method also includes determining, by the processing system, based at least in part on a model trained to identify the output of the speaker device, that the additional audio signal corresponds to an utterance of a user. The method further includes initiating a reduction in an audio output level of the speaker device based on determining that the additional audio signal corresponds to the utterance of the user.

Abstract: The present technology relates to an information processing apparatus, an information processing method, and a program that enable provision of information to a user while protecting privacy. An extraction unit that extracts information from an utterance of a user, an inquiry unit that makes an inquiry to another apparatus when a request from the user is given, and a supplementation unit that supplements the information extracted by the extraction unit to inquiry content when the inquiry unit makes an inquiry are provided. A determination unit that determines whether or not the information supplemented by the supplementation unit is information regarding privacy is further provided. The information extracted by the extraction unit is registered to a database in association with a flag indicating whether or not the information is the information regarding privacy. The present technology can be applied to an information processing apparatus that presents information to a user.

Abstract: Techniques for capturing spoken user inputs while a device is prevented from capturing such spoken user inputs are described. When a first device becomes incapable of capturing spoken user inputs intended for a system, a second device, for capturing such spoken user inputs, may be identified. The second device may be identified based on the second device being connected to a same vehicle computing system as the first device. The second device may be enabled to capture spoken user inputs, intended for the system, until the first device is again able to capture such spoken user inputs.

Abstract: Conversational systems are required to be capable of handling more sophisticated interactions than providing factual answers only. Such interactions are handled by resolving abstract anaphoric references in conversational systems which includes antecedent fact references and posterior fact references. The present disclosure resolves abstract anaphoric references in conversational systems using hierarchically stacked neural networks. In the present disclosure, a deep hierarchical maxpool network based model is used to obtain a representation of each utterance received from users and a representation of one or more generated sequences of utterances. The obtained representations are further used to identify contextual dependencies with in the one or more generated sequences which helps in resolving abstract anaphoric references in conversational systems.

Abstract: Techniques are provided for reverberation compensation for far-field speaker recognition. A methodology implementing the techniques according to an embodiment includes receiving an authentication audio signal associated with speech of a user and extracting features from the authentication audio signal. The method also includes scoring results of application of one or more speaker models to the extracted features. Each of the speaker models is trained based on a training audio signal processed by a reverberation simulator to simulate selected far-field environmental effects to be associated with that speaker model. The method further includes selecting one of the speaker models, based on the score, and mapping the selected speaker model to a known speaker identification or label that is associated with the user.

Abstract: A computer-implemented method is presented for implementing multi-aspect sentiment analysis by collaborative attention allocation. The method includes extracting a sequence of word vectors from a sentence received from a data stream, feeding the sequence of word vectors to long short-term memory (LSTM) neural networks to generate a sequence of hidden states corresponding to the sequence of word vectors, generating a plurality of aspect embedding vectors for each aspect, employing an attention mechanism to determine attention weight vectors concurrently for all aspects, and outputting predicted sentiments for each aspect of the sentence to a user interface of a computing device.

Abstract: Disclosed are various embodiments for detecting replay attacks in voice-based authentication systems. In one embodiment, audio is captured via an audio input device. It is then verified that the audio includes a voice authentication factor spoken by a user. If it is determined that the audio includes unexpected environmental audio in addition to the voice authentication factor that has been verified, one or more actions may be performed.

Abstract: The technology disclosed herein may determine timing windows for speech captions of an audio stream. In one example, the technology may involve accessing audio data comprising a plurality of segments; determining, by a processing device, that one or more of the plurality of segments comprise speech sounds; identifying a time duration for the speech sounds; and providing a user interface element corresponding to the time duration for the speech sounds, wherein the user interface element indicates an estimate of a beginning and ending of the speech sounds and is configured to receive caption text associated with the speech sounds of the audio data.

Abstract: A computerized method of assessing a chatbot conversation includes: extracting one or more messages from the conversation; determining, based on the one or more messages, an existing business opportunity value score using a business opportunity state detector module; determining, based on the one or more messages, an existing user experience score using a user experience state detector module of the computing device; determining, based on the one or more messages, a future business opportunity value score using a future business opportunity predictor module of the computing device; determining, based on the one or more messages, a future user experience score using the future user experience predictor module of the computing device; calculating a composite score indicating whether human intervention in the chatbot conversation is desirable; and generating a display signal including a status indicator, for review by a human agent, reflecting a desirability of human intervention in the chatbot conversation.

Abstract: A method and system for changing content of a window of an application program is provided. A contextual window system displays a window with content based on a current context of the window. The contextual window system receives from a user a context string for a new context for the window. When the context string includes a command, the contextual window system performs a function of the application program that implements the command to change from the current context of the window to the new context of the window. When the context string does not specify a command, the contextual window system submits the context string as a query for data of the application program to change from the current context of the window to the new context of the window. The contextual window system then modifies the content of the window to reflect the new context of the window.

Abstract: Systems including a universal context aggregator configured to pre-fetch context information that may be used to perform various processes with respect to a user input are described. The aggregator may have access to data representing what context information components of the system routinely request in various situations. When a particular situation is present, prior to being queried, the aggregator may pre-fetch context information that the aggregator is likely to be queried for.

Abstract: A dialysis system, comprising: a dialysis machine; a voice recognition component configured to identify a voice command in audio information received by a microphone of the dialysis system; an authentication component configured to determine a source of the voice command; and a processor configured to perform a function determined based on the voice command.

Abstract: A computer-implemented method comprising: receiving, by a computing device, an input phrase from a text generator; determining, by the computing device, a complexity level for an audience; generating, by the computing device, a plurality of target phrases including a modification of the input phrase; generating, by the computing device, respective readability scores for each of the plurality of target phrases; mapping, by the computing device, the plurality of the target phrases to the target audience complexity level to select a particular target phrase of the plurality of the target phrases; and outputting, by the computing device, the selected particular target phrase to a text-to-speech (T2S) component to cause the T2S component to output the selected particular target phrase as audible speech.

Abstract: A method, computer system, and a computer program product for voice activated inventory management is provided. The present invention may include recording an audio feed of a customer product query from a customer and a staff response from a staff member. The present invention may then include identifying a product requested by the customer. The present invention may also include identifying an inventory status in the staff response. The present invention may also include determining that a negative inventory status trigger is detected in the identified inventory status associated with the identified product requested by the customer. The present invention may further include, in response to determining that the negative inventory status trigger is detected in the identified inventory status associated with the identified product requested by the customer, storing, in an inventory database, a plurality of customer query data associated with the identified product requested by the customer.

Abstract: This disclosure generally relates to a system for communicating data generated by a wearable device to one or more server devices for analysis. The one or more server devices may transmit activity level data, or a graphical representation thereof, for a wearer of a wearable device to a device associated with a healthcare provider. The activity level data may include one or more of an active minutes element, a television time element, a word count element, a sleep duration element and a reading duration and score element.

Abstract: Devices and methods are provided for extracting content from audio files. The device may determine starting and ending quotation marks in a text file, and a string between the starting and ending quotation marks. The device may determine that a verb is near the starting quotation mark or the ending quotation mark. The device may determine, based on the verb, that the string is attributed to a character name in the text file. The device may determine a first time in a first audio file including an audio representation of the text file, and may determine a second time in the first audio file, wherein the first time is before the first word and the second time is after the second word. The device may generate a second audio file by extracting audio from the first audio file based on the first and second times.

Abstract: A vocal feedback device comprising: a microphone; a fundamental frequency accentuator electrically connected to the microphone, a delay circuit electrically connected to the fundamental frequency accentuator, and a speaker electrically connected to the delay circuit. The device configured to convert vocal utterances received at the microphone into an electrical signal, impose a time delay before transmitting the electrical signal, after the time delay, transmit the electrical signal to the speaker, and convert the electrical signal to an audio signal using the speaker, the audio signal being a replication of the vocal utterances.