Abstract: A method of communication includes determining, at a mobile device, a speech quality metric for an incoming speech signal associated with a voice call. The speech quality metric is based on an environment of the mobile device. The method also includes converting incoming speech associated with the incoming speech signal to text in response to a determination that the speech quality metric fails to satisfy a speech quality metric threshold. The method further includes displaying the text at a display screen of the mobile device during the voice call.

Abstract: A voice to text model used by a voice-enabled electronic device is dynamically and in a context-sensitive manner updated to facilitate recognition of entities that potentially may be spoken by a user in a voice input directed to the voice-enabled electronic device. The dynamic update to the voice to text model may be performed, for example, based upon processing of a first portion of a voice input, e.g., based upon detection of a particular type of voice action, and may be targeted to facilitate the recognition of entities that may occur in a later portion of the same voice input, e.g., entities that are particularly relevant to one or more parameters associated with a detected type of voice action.

Abstract: A log spectral envelope sequence L0, L1, . . . , LN?1 and an envelope code for the log spectral envelope sequence L0, L1, . . . , LN?1 are obtained. The log spectral envelope sequence L0, L1, . . . , LN?1 is an integer value sequence corresponding to binary logarithms of respective sample values of a spectral envelope sequence and is an integer value sequence whose total sum is 0. For a quantized spectral sequence {circumflex over (?)}X0, {circumflex over (?)}X1, . . . , {circumflex over (?)}XN?1, a smoothed spectral sequence ˜X0, ˜X1, . . . , ˜XN?1 is obtained by: for {circumflex over (?)}Xk with Lk being a positive value, adopting {circumflex over (?)}Xk with Lk digits from its least significant digit removed as ˜Xk; for {circumflex over (?)}Xk with Lk being a negative value, adopting {circumflex over (?)}Xk with ?Lk digits added to its least significant digit in accordance with a predefined rule as ˜Xk; and when Lk is 0, adopting {circumflex over (?)}Xk as ˜Xk.

Abstract: A system, method, and wireless earpieces for implementing a virtual assistant. A request is received from a user to be implemented by wireless earpieces. A virtual assistant is executed on the wireless earpieces. An action is implemented to fulfill the request utilizing the virtual assistant. The wireless earpieces may be a set of wireless earpieces and the virtual assistant may be implemented independently by the wireless earpieces.

Abstract: A method of speaker identification, comprises: receiving an audio signal representing speech; removing effects of a channel and/or noise from the received audio signal to obtain a cleaned audio signal; obtaining an average spectrum of at least a part of the cleaned audio signal; and comparing the average spectrum with a long term average speaker model for an enrolled speaker. Based on the comparison, it can be determined whether the speech is the speech of the enrolled speaker.

Abstract: A device for detecting a morpheme tagging corpus error, of the present invention, includes: an attribute generating unit for generating attributes for word phrases included in an input corpus, by using a kernel to which a rough set theory is applied; and an attribute statistics processing unit for generating part-of-speech tagging corpus error data through the calculation of attributes and frequency count for the same word phrases by counting attributes for the same word phrase among the word phrases, and thus the present invention can detect, quantify, and modify errors included in a corpus (learning data) required in learning for classifier generation and recognition for natural language processing.

Abstract: Techniques are disclosed relating to determining phonetic similarity using machine learning. The techniques include accessing training data that includes a first set of words of a native language and a second set of words corresponding to verified transliterations of the first set of words from the native language to a target language. Further, they include generating a set of new transliterations of the first set of words from the native language to the target language and storing comparison information based on a comparison between words from the second set of words and word from the set of new transliterations of the first set of words. Finally, a similarity score is determined between a first word of the target language and a second word of the target language based on the comparison information.

Abstract: An interactive method and a device thereof are provided. The method includes obtaining voice data of the object in response to determining that the object is facing the interactive device and is in the utterance state; and establishing an interaction between the object and the interactive device based on the voice data. The method solves the technical problems in which current interactions need to set up wakeup terms for interactive devices which are prone to false wakeups through the wakeup terms due to an existence of a relatively small number of wakeup terms. The above methods can implement the technical effects of remote interactions without the need of a wakeup term.

Abstract: Described herein is a system for rescoring automatic speech recognition hypotheses for conversational devices that have multi-turn dialogs with a user. The system leverages dialog context by incorporating data related to past user utterances and data related to the system generated response corresponding to the past user utterance. Incorporation of this data improves recognition of a particular user utterance within the dialog.

Abstract: This disclosure relates to a mechanism to create conversational agents from API specifications based on domain-specific inputs. The conversational agents may provide the functionalities exposed by the underlying API to users engaging with the conversational agent. Thus, the user may execute actions exposed by the API specification using natural language in a conversational, comfortable, and familiar fashion.

Abstract: The present invention allows for the capture and sentiment analysis of text the customer inputs into a chat, but never actually sends to the customer service representative (ghost text). The system captures this ghost text with a ghost capture system (GCS) software module. The GCS module analyzes the ghost text to generate metadata. The ghost text and metadata are used by a sentiment analysis engine to apply appropriate sentiment to the ghost text. The sentiment and ghost text are routed to a customer service representative (CSR). This provides the customer service agent with additional detail and information about a customer's emotions during a text chat conversation, allowing the CSR to determine a court of interaction not only based on the customer's response, but also based on the ghost text and the sentiment from the ghost text.

Abstract: The present disclosure proposes a speech processing method and a cloud-based speech processing apparatus. The speech processing method includes: acquiring a piece of speech to be recognized collected by a terminal; performing a speech recognition on the piece of speech to be recognized; detecting whether the piece of speech to be recognized ends during the speech recognition; and feeding back a recognized result of the piece of speech to be recognized to the terminal when it is detected that the piece of speech to be recognized ends.

Abstract: The present teaching relates to method, system, medium, and implementations for speech recognition. An audio signal is received that represents a speech of a user engaged in a dialogue. A visual signal is received that captures the user uttering the speech. A first speech recognition result is obtained by performing audio based speech recognition based on the audio signal. Based on the visual signal, lip movement of the user is detected and a second speech recognition result is obtained by performing lip reading based speech recognition. The first and the second speech recognition results are then integrated to generate an integrated speech recognition result.

Abstract: A computer implemented method of controlling the incidence of spoilers in a conversation that includes calculating a baseline of events of interest for participants to a conversation, the calculating of the baseline including machine learning applied to interest indicators for the participants, the interest indicators being derived from data collected from social media accounts of the participants and calendars of the participants in response the participant granting permission for the data collection. The method further includes monitoring of real time conversation between the parties for keywords indicative of a topic of the real time conversation, wherein a spoiler message is predicted when the keywords substantially match the baseline for the events of interest. The method may further include sending an anti-spoiler signal to the participants of the conversation when the keywords substantially match the baseline for the events of interest.

Abstract: Methods, computer program products, and systems are presented. The method computer program products, and systems can include, for instance: obtaining vocal utterance data representing vocal utterances of multiple users within a venue; processing the vocal utterance data to return metadata associated to the vocal utterance data; predicting using the metadata an item for acquisition by one or more user of the multiple users; and returning an action decision in dependence on the predicting.

Abstract: Data associated with a selectively offline capable voice action is locally persisted in a voice-enabled electronic device whenever such an action cannot be competed locally due to the device being offline to enable the action to later be completed after online connectivity has been restored. Synchronization with an online service and/or another electronic device, and/or retrieval of context sensitive data from an online service may be performed after online connectivity has been restored to enable the voice action to thereafter be completed.

Abstract: Described herein are systems and methods for a general, scalable, end-to-end framework that uses a generative adversarial network (GAN) objective to enable robust speech recognition. Encoders trained with the proposed approach enjoy improved invariance by learning to map noisy audio to the same embedding space as that of clean audio. Embodiments of a Wasserstein GAN framework increase the robustness of seq-to-seq models in a scalable, end-to-end fashion. In one or more embodiments, an encoder component is treated as the generator of GAN and is trained to produce indistinguishable embeddings between labeled and unlabeled audio samples. This new robust training approach can learn to induce robustness without alignment or complicated inference pipeline and even where augmentation of audio data is not possible.

Abstract: Techniques for joining a device of a third user to a communication between a device of a first user and a device of a second user are described herein. For instance, two or more users may utilize respective computing devices to engage in a telephone call, a video call, an instant-messaging session, or any other type of communication in which the users communicate with each other audibly and/or visually. In some instances, a first user of the two users may issue a voice command requesting to join a device of a third user to the communication. One or more computing devices may recognize this voice command and may attempt to join a device of a third user to the communication.

Abstract: Systems and methods are disclosed for enabling verbal interaction with an NLUI application without relying on express wake terms. The NLUI application receives an audio input comprising a plurality of terms. In response to determining that none of the terms is an express wake term pre-programmed into the NLUI application, the NLUI application determines a topic for the plurality of terms. The NLUI application then determines whether the topic is within a plurality of topics for which a response should be generated. If the determined topic of the audio input is within the plurality of topics, the NLUI application generates a response to the audio input.

Abstract: An input method editor (IME) is associated with a local user. Memory stores local data and a processor, coupled to the memory, is configured to receive input from a local, first user, obtain shared data associated with at least a remote, second user from a remote server and generate prediction candidates and conversion candidates based on the input provided by the local, first user and correlation of the input and the obtained shared data.