Abstract: An audio processor for generating a frequency enhanced audio signal from a source audio signal has: an envelope determiner for determining a temporal envelope of at least a portion of the source audio signal; an analyzer for analyzing the temporal envelope to determine temporal values of certain features of the temporal envelope; a signal synthesizer for generating a synthesis signal, the generating having placing pulses in relation to the determined temporal values, wherein the pulses are weighted using weights derived from amplitudes of the temporal envelope related to the temporal values, where the pulses are placed; and a combiner for combining at least a band of the synthesis signal that is not included in the source audio signal and the source audio signal to obtain the frequency enhanced audio signal.

Abstract: A key word detection apparatus and a method for low-power voice-activated devices are presented. A first signal processing module operates with a first transducer to receive an incoming signal and generates a first sample. A second signal processing module operates with a second transducer which receives an incoming signal and generates a second sample. In summary, a signal processing system, in particular a key word detection system, has a first low power module that wakes up a second higher power module. The second module uses signals from the first module in order to improve accuracy of key word detection or other signal processing tasks.

Abstract: Systems described herein may use machine classifiers to perform a variety of natural language understanding tasks including, but not limited to multi-turn dialogue generation. Machine classifiers in accordance with aspects of the disclosure may model multi-turn dialogue as a one-to-many prediction task. The machine classifier may be trained using adversarial bootstrapping between a generator and a discriminator with multi-turn capabilities. The machine classifiers may be trained in both auto-regressive and traditional teacher-forcing modes, with the maximum likelihood loss of the auto-regressive outputs being weighted by the score from a metric-based discriminator model. The discriminators input may include a mixture of ground truth labels, the teacher-forcing outputs of the generator, and/or negative examples from the dataset. This mixture of input may allow for richer feedback on the autoregressive outputs of the generator.

Abstract: Certain aspects of the present disclosure provide a method for performing voice activity detection, including: receiving audio data from an audio source of an electronic device; generating a plurality of model input features using a hardware-based feature generator based on the received audio data; providing the plurality of model input features to a hardware-based voice activity detection model; receiving an output value from the hardware-based voice activity detection model; and determining a presence of voice activity in the audio data based on the output value.

Abstract: Techniques are provided for evaluating multiple machine learning models to identify issues with a communication. One method comprises applying an audio signal associated with a communication to at least two of: (i) a trigger word analysis module that evaluates contextual information to determine if a trigger word is detected in the audio signal; (ii) an audio activity pattern analysis module that determines if a silence pattern anomaly is detected; and (iii) a communication application analysis module that evaluates features provided by a communication application relative to applicable thresholds; and combining results of the at least two of the trigger word analysis module, the audio activity pattern analysis module and the communication application analysis module to identify a communication issue. The combining may evaluate an accuracy of the trigger word analysis module, the audio activity pattern analysis module and/or the communication application analysis module to combine the results.

Abstract: A system configured to perform utterance detection using data processing techniques that are similar to those used for object detection is provided. For example, the system may treat utterances within audio data as analogous to an object represented within an image and employ techniques to separate and identify individual utterances. The system may include one or more trained models that are trained to perform utterance detection. For example, the system may include a first module to process input audio data and identify whether speech is represented in the input audio data, a second module to apply convolution filters, and a third module configured to determine a boundary identifying a beginning and ending of a portion of the input audio data along with an utterance score indicating how closely the portion of the input audio data represents an utterance.

Abstract: A computer-implemented method in accordance with one embodiment includes, in response to a submission of an input dataset to an artificially intelligent application, receiving an explanation from each module of the application. The modules are configured within the application in a serial sequence in which each module, upon receiving the input dataset and any input generated by an immediately preceding module of the serial sequence, generates output that is forwarded as input to a next module, if any, in the sequence. A determination is made that at least two of the received explanations are semantically inconsistent.

Abstract: Systems, methods, and computer programming products for alleviating ambiguity amongst the terms and language displayed by the user interface of software products and services. The disclosed solutions catalog terms displayed by the UI of software and services and identify where overlapping terms with the same or substantially similar term names are presented by the UI but have different meanings than the software most familiar to the user. Natural language processing is leveraged to derive meanings of software terms using the context of the surrounding words and text elements within the UI, as well as product documentation, error messages, sentiment and other textual clues. Ambiguity among overlapping terms is alleviated by modifying the UI, highlighting differences in term definitions from the software or services a user is most familiar with using, and updating the UI in a manner that differentiates the overlapping terms displayed by accessed products or services.

Abstract: The present invention relates to a voice conversation analysis apparatus and a method therefor and, more specifically, to: a voice conversation analysis apparatus categorizing voices generated during a voice conversation so as to predict required functions and further analyzing the voices so as to provide proper functions; and a method therefor. In addition, disclosed are: an artificial intelligence (AI) system for simulating the functions of recognition, decision-making, and the like of the human brain by using a machine learning algorithm; and an application thereof.

Abstract: Natural language processing systems and methods are disclosed herein. In some embodiments, digital document information comprising text is received. The digital document information may be processed through word and character encoding operations to generate word and character vectors while retaining document location information for the words and characters. The data may be then be processed by a series of convolution and maximum pooling operations to obtain maximum valued elements from the data. The document location information as well as the maximum values element data may be further processed for semantic classification of the data using a semantic classifier and bounding box regression.

Abstract: The presentation of an automated assistant response may be selectively pre-empted in response to a hot-word free utterance that is received during the presentation and that is determined to be likely directed to the automated assistant. The determination that the utterance is likely directed to the automated assistant may be performed, for example, using an utterance classification operation that is performed on audio data received during presentation of the response, and based upon such a determination, the response may be pre-empted with another response associated with the later-received utterance. In addition, the duration that is used to determine when a session should be terminated at the conclusion of a conversation between a user and an automated assistant may be dynamically controlled based upon when the presentation of a response has completed.

Abstract: A computer-implemented sound signal generation method includes: obtaining a first sound source spectrum of a sound signal to be generated; obtaining a first spectral envelope of the sound signal; and estimating fragment data representative of samples of the sound signal based on the obtained first sound source spectrum and the obtained first spectral envelope.

Abstract: A conversational interface generation method, system, and computer program product that includes determining a conversational artifact for a computer program from a specification of the computer program and generating a conversational interface for the computer program based on the conversational artifact for the computer program included in the specification.

Abstract: Embodiments described herein provide methods and systems for training task-oriented dialogue (TOD) language models. In some embodiments, a TOD language model may receive a TOD dataset including a plurality of dialogues and a model input sequence may be generated from the dialogues using a first token prefixed to each user utterance and a second token prefixed to each system response of the dialogues. In some embodiments, the first token or the second token may be randomly replaced with a mask token to generate a masked training sequence and a masked language modeling (MLM) loss may be computed using the masked training sequence. In some embodiments, the TOD language model may be updated based on the MLM loss.

Abstract: An electronic apparatus, including a memory configured to store a first artificial intelligence model; and a processor connected to the memory and configured to: based on receiving an input audio signal, obtain an input frequency spectrum image representing a frequency spectrum of the input audio signal, input the input frequency spectrum image to the first artificial intelligence model, obtain an output frequency spectrum image from the first artificial intelligence model, obtain an output audio signal based on the output frequency spectrum image, wherein the first artificial intelligence model is trained based on a target learning image, and wherein the target learning image represents a target frequency spectrum of a specific style, and is obtained from a second artificial intelligence model based on a random value.

Abstract: Systems, methods, and computer-readable storage media for responding to a query using a neural network and natural language processing. If necessary, the system can request disambiguation, then parse the query using a trained machine-learning classifier, resulting in at least one of an identified subject or an identified domain of the text query. The system can determine if the user is authorized to retrieve answers to the query and, if so, retrieve factual data associated with the query. The system can then retrieve a response template, and fill in the template with the retrieved facts. The system can then determine, by executing a machine comprehension model on the filled response template, a probable readability token, portion of text, of at least a portion of the filled response template and, upon identifying that the probable readability is above a threshold, reply to the text query with the at least a portion of the filled response template.

Abstract: Techniques for automatically detecting when a ride requester has requested a ride-share ride on behalf of a guest rider using some or all of the communications between the driver and ride requester are described herein. For example, a server can obtain chat logs between a ride requester and a driver and process the chat logs to identify whether the ride requester has requested a ride on behalf of a guest rider. In particular, the server can train an artificial intelligence model (e.g., a machine learning model) to predict potential guest rider behavior. Once trained, the server can obtain chat logs comprising chat messages sent between a driver and a ride requester, and apply a representation of the chat logs as an input to the trained artificial intelligence model to determine whether guest rider behavior is detected.

Abstract: An automatic speech recognition system and a method thereof are provided. The system includes an encoder and a decoder. The encoder comprises a plurality of encoder layers. At least one encoder layer includes a plurality of encoder sublayers fused into one or more encoder kernels. The system further comprises a first pair of ping-pong buffers communicating with the one or more encoder kernels. The decoder comprises a plurality of decoder layers. At least one decoder layer includes a plurality of decoder sublayers fused into one or more decoder kernels. The decoder receives a decoder output related to the encoder output and generates a decoder output. The encoder sends the decoder output to a beam search kernel.

Abstract: Methods, systems, and computer program products for automatically modifying responses from generative models using artificial intelligence techniques are provided herein. A computer-implemented method includes obtaining data pertaining to at least one conversation involving at least one automated conversation exchange software program and at least one user; identifying, among words proposed by the at least one automated conversation exchange software program in connection with the at least one conversation, words qualifying as belonging to one or more predetermined categories by processing the obtained data using artificial intelligence techniques; determining, by processing the identified words and at least one word-based data source, one or more alternate words; modifying at least a portion of the proposed words by replacing at least a portion of the identified words with at least a portion of the one or more alternate words; and performing at least one automated action based on the modifying.

Abstract: A method includes obtaining, by an electronic device, an audio segment comprising one or more audio events of a target subject. The method also includes extracting, by the electronic device, audio embeddings from the one or more audio events using an embedding model, the embedding model comprising a trained machine learning model. The method further includes comparing, by the electronic device, the extracted audio embeddings with a match profile of the target subject, the match profile generated during an enrollment stage. The method also includes generating, by the electronic device, a label for the audio segment based on whether or not the extracted audio embeddings match the match profile, wherein the label enables correlation of the audio segment with the target subject for monitoring a health condition of the target subject.