Abstract: Systems and methods for correcting a voice query based on a subsequent voice query with a lower pronunciation rate. In some aspects, the systems and methods calculate first and second pronunciation rates of first and second voice queries. The systems and methods determine that the second pronunciation rate is lower than the first pronunciation rate and determine a first candidate pronunciation time for a first candidate word from the first voice query. The systems and methods determine a second candidate pronunciation time, adjusted to the first pronunciation rate, for the second candidate word from the second voice query. The systems and methods determine that the first candidate pronunciation time matches the second candidate pronunciation time and generate a third voice query based on the first voice query by replacing the first candidate word with the second candidate word.

Abstract: Systems and methods for dynamic sequence-based adjustment of prompt generation are provided. The system can receive a first interaction and a second interaction via a client device and identify a first sequence based on the first interaction and the second interaction. The system can map the first sequence to a node data structure and identify a node in the node data structure that matches the first sequence. The system can generate an adjusted parameter for a first digital component object responsive to a match with an attribute of the node in the node data structure. The system can execute a real-time digital component selection process among a plurality of digital component objects including the first digital component object to select the first digital component object. The system can transmit a prompt with the first digital component object to a client device to cause the client device to present the prompt.

Abstract: A voice processing method, a voice communication device, and a computer program product thereof are disclosed. The method comprises the steps of: receiving a transmitting voice signal from a receiver end communication device; determining a frequency range of the transmitting voice signal; receiving an original voice signal from a first user; processing the original voice signal to a processed voice signal, wherein the processed voice signal is generated based on the frequency range of the transmitting voice signal; and outputting the processed voice signal to the receiver end communication device.

Abstract: During text-to-speech processing, a speech model creates output audio data, including speech, that corresponds to input text data that includes a representation of the speech. A spectrogram estimator estimates a frequency spectrogram of the speech; the corresponding frequency-spectrogram data is used to condition the speech model. A plurality of acoustic features corresponding to different segments of the input text data, such as phonemes, syllable-level features, and/or word-level features, may be separately encoded into context vectors; the spectrogram estimator uses these separate context vectors to create the frequency spectrogram.

Abstract: In one aspect, a device includes a processor, a microphone accessible to the processor, at least a first sensor that is accessible to the processor, and storage accessible to the processor. The storage bears instructions executable by the processor to receive first input from the microphone that is generated based on audible input from a user. The instructions are also executable by the processor to receive second input from the first sensor, perform natural language understanding based on the first input, augment the natural language understanding based on the second input, and provide an output based on the augmentation.

Abstract: A virtual vocabulary database is provided for use with a with a particular user database as part of a speech recognition system. Vocabulary elements within the virtual database are imported from the user database and are tagged to include numerical data corresponding to the historical use of the vocabulary element within the user database. For each speech input, potential vocabulary element matches from the speech recognition system are provided to the virtual database software which creates virtual sub-vocabularies from the criteria according to predefined criteria templates. The software then applies vocabulary element weighting adjustments according to the virtual sub-vocabulary weightings and applies the adjustment to the default weighting provided by the speech recognition system. The modified weightings are returned with the associated vocabulary elements to the speech engine for selection of an appropriate match to the input speech.

Abstract: Techniques for disambiguating which profile, of multiple profiles, is to be used to respond to a user input are described. A device located in a communal space (e.g., a hotel room or suite of rooms, conference room, hospital room, etc.) may be associated with a device profile and a user profile of a user presently occupying the communal space. When the user inputs a command to the device (either by text or speech), a system associated with the device determines the profiles (e.g., a device profile and a user profile) associated with the device. The system determines one or more policies associated with the device. The one or more policies may correspond to rules for disambiguating which profile to use to execute with respect to the user input. Using the one or more policies, the system determines which profile is to be used, and causes a speechlet component to execute using information specific to the determined profile.

Abstract: A method, computer system, and a computer program product for optimizing speech recognition in a smart medical room. The present invention may include selecting, from a database, one or more speech domain models based on a plurality of signals from a plurality of biometric sensors associated with a plurality of medical equipment, wherein the one or more speech domain models are trained with one or more feedback from a clinician based on a medical encounter and from a continuous feedback display in the smart medical room, wherein the one or more feedback from the clinician is based on an optional notification to the clinician to confirm the one or more speech models in use.

Abstract: Speech or non-speech detection techniques are discussed and include updating a speech pattern model using probability scores from an acoustic model to generate a score for each state of the speech pattern model, such that the speech pattern model includes a first non-speech state having multiple self loops each associated with a non-speech probability score of the probability scores, a plurality of speech states following the first non-speech state, and a second non-speech state following the speech states, and detecting speech based on a comparison of a score of the first non-speech state and a score of the last speech state of the multiple speech states.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for hotword detection on multiple devices are disclosed. In one aspect, a method includes the actions of receiving audio data that corresponds to an utterance. The actions further include determining that the utterance likely includes a particular, predefined hotword. The actions further include transmitting (i) data indicating that the computing device likely received the particular, predefined hotword, (ii) data identifying the computing device, and (iii) data identifying a group of nearby computing devices that includes the computing device. The actions further include receiving an instruction to commence speech recognition processing on the audio data. The actions further include in response to receiving the instruction to commence speech recognition processing on the audio data, processing at least a portion of the audio data using an automated speech recognizer on the computing device.

Abstract: A computer-implemented method, system, and computer program product are provided for content aware heterogeneous log pattern comparative analysis. The method includes receiving, by a processor-device, a plurality of heterogeneous logs. The method also includes extracting, by the processor-device, a plurality of log syntactic patterns from the plurality of heterogenous logs. The method additionally includes generating, by the processor-device, latent representation vectors for each of the plurality of log syntactic patterns. The method further includes predicting, by the processor-device, an anomaly from the clustered latent representation vectors. The method also includes controlling an operation of a processor-based machine to react in accordance with the anomaly.

Abstract: A speech model includes a sub-model corresponding to a vocal attribute. The speech model generates an output waveform using a sample model, which receives text data, and a conditioning model, which receives text metadata and produces a prosody output for use by the sample model. If, during training or runtime, a different vocal attribute is desired or needed, the sub-model is re-trained or switched to a different sub-model corresponding to the different vocal attribute.

Abstract: A decoder for generating an audio output signal having one or more audio output channels is provided, having a receiving interface for receiving an audio input signal having a plurality of audio object signals, for receiving loudness information on the audio object signals, and for receiving rendering information indicating whether one or more of the audio object signals shall be amplified or attenuated, further having a signal processor for generating the one or more audio output channels of the audio output signal, configured to determine a loudness compensation value depending on the loudness information and depending on the rendering information, and configured to generate the one or more audio output channels of the audio output signal from the audio input signal depending on the rendering information and depending on the loudness compensation value. One or more by-pass audio object signals are employed for generating the audio output signal. Moreover, an encoder is provided.

Abstract: A speech model is trained using multi-task learning. A first task may correspond to how well predicted audio matches training audio; a second task may correspond to a metric of perceived audio quality. The speech model may include, during training, layers related to the second task that are discarded at runtime.

Abstract: An electronic device and an method of the electronic device are provided, where the electronic device maintains a context that does not reflect a request for a secret conversation, in response to the request for the secret conversation being received from a first user, and generates a response signal to a voice signal of a second user based on the maintained context, in response to an end of the secret conversation with the first user.

Abstract: A method, computer system, and a computer program product for altering a written communication based on a dress style associated with a recipient is provided. The present invention may include receiving a plurality of visual data associated with the recipient. The present invention may also include analyzing the received plurality of visual data. The present invention may then include determining the dress style associated with the recipient based on the analyzed plurality of visual data. The present invention may further include retrieving a writing style associated with the recipient from a knowledge base based on the determined dress style. The present invention may also include generating a plurality of writing guidelines based on the retrieved writing style associated with the recipient.

Abstract: Systems and methods for utilizing microphone array information for acoustic modeling are disclosed. Audio data may be received from a device having a microphone array configuration. Microphone configuration data may also be received that indicates the configuration of the microphone array. The microphone configuration data may be utilized as an input vector to an acoustic model, along with the audio data, to generate phoneme data. Additionally, the microphone configuration data may be utilized to train and/or generate acoustic models, select an acoustic model to perform speech recognition with, and/or to improve trigger sound detection.

Abstract: A method, computer system, and a computer program product for altering a written communication based on a dress style associated with a recipient is provided. The present invention may include receiving a plurality of visual data associated with the recipient. The present invention may also include analyzing the received plurality of visual data. The present invention may then include determining the dress style associated with the recipient based on the analyzed plurality of visual data. The present invention may further include retrieving a writing style associated with the recipient from a knowledge base based on the determined dress style. The present invention may also include generating a plurality of writing guidelines based on the retrieved writing style associated with the recipient.

Abstract: Voice activity detection (VAD) is an enabling technology for a variety of speech based applications. Herein disclosed is a robust VAD algorithm that is also language independent. Rather than classifying short segments of the audio as either “speech” or “silence”, the VAD as disclosed herein employees a soft-decision mechanism. The VAD outputs a speech-presence probability, which is based on a variety of characteristics.

Abstract: Systems and methods for detecting, classifying, and correcting acoustic (waveform) events are provided. In one example embodiment, a computer-implemented method includes obtaining, by a computing system, audio data from a source. The method includes accessing, by the computing system, data indicative of a machine-learned acoustic detection model. The method includes inputting, by the computing system, the audio data from the source into the machine-learned acoustic detection model. The method includes obtaining, by the computing system, an output from the machine-learned acoustic detection model. The output is indicative of an acoustic event associated with the source. The method includes providing, by the computing system, data indicative of a notification to a user device. The notification indicates the acoustic event and response(s) for selection by a user.