Abstract: An apparatus for speech model with personalization via ambient context harvesting, is described herein. The apparatus includes a microphone, context harvesting module, confidence module, and training module. The context harvesting module is to determine a context associated with the captured audio signals. A confidence module is to determine a confidence of the context as applied to the audio signals. A training module is to train a neural network in response to the confidence being above a predetermined threshold.

Abstract: An apparatus for speech model with personalization via ambient context harvesting, is described herein. The apparatus includes a microphone, context harvesting module, confidence module, and training module. The context harvesting module is to determine a context associated with the captured audio signals. A confidence module is to determine a confidence of the context as applied to the audio signals. A training module is to train a neural network in response to the confidence being above a predetermined threshold.

Abstract: A student model may be trained in two stages by using two teacher models, respectively. The first teacher model has been trained with a pretraining dataset. The second teacher model has been trained with a training dataset that is specific to a task to be performed by the student model. In the first stage, the student model may be generated based on a structure of the first teacher model. Internal parameters of the student model are adjusted through a pretraining process based on the first teacher model and the pretraining dataset. Weights of the student model may be pruned during the pretraining process. In the second stage, a sparsity mask is generated for the student model to lock the sparsity pattern generated from the first stage. Further, some of the internal parameters of the student model are modified based on the second teacher model and the training dataset.

Abstract: Methods and apparatus for a knowledge-based deep learning refactoring model with tightly integrated functional nonparametric memory are disclosed. An example non-transitory computer readable medium comprises instructions that, when executed, cause a machine to at least estimate a first information extraction cost corresponding to retrieval of information from a local knowledge base, estimate a second information extraction cost corresponding retrieval of information from a remote knowledge base, select an information source based on the first and second estimated information extraction costs, query the selected information source, in response to determining that the selected information source was an external information source, store the queried information in the local knowledge base, organize the stored information in the local knowledge base, and return the queried information.

Abstract: An apparatus for speech model with personalization via ambient context harvesting, is described herein. The apparatus includes a microphone, context harvesting module, confidence module, and training module. The context harvesting module is to determine a context associated with the captured audio signals. A confidence module is to determine a confidence of the context as applied to the audio signals. A training module is to train a neural network in response to the confidence being above a predetermined threshold.

Abstract: Technologies for detecting an end of a sentence in automatic speech recognition are disclosed. An automatic speech recognition device may acquire speech data, and identify phonemes and words of the speech data. The automatic speech recognition device may perform a syntactic parse based on the recognized words, and determine an end of a sentence based on the syntactic parse. For example, if the syntactic parse indicates that a certain set of consecutive recognized words form a syntactically complete and correct sentence, the automatic speech recognition device may determine that there is an end of a sentence at the end of that set of words.

Abstract: A system and method for syntactic re-ranking of possible transcriptions generated by automatic speech recognition are disclosed. A computer system accesses acoustic data for a recorded spoken language and generates a plurality of potential transcriptions for the acoustic data. The computer system scores the plurality of potential transcriptions to create an initial likelihood score for the plurality of potential transcriptions. For a particular potential transcription in the plurality of transcriptions, the computer system generates a syntactical likelihood score. The computer system creates an adjusted score for the particular potential transcription by combining the initial likelihood score and the syntactic likelihood score for the particular potential transcription.

Abstract: A system and method for syntactic re-ranking of possible transcriptions generated by automatic speech recognition are disclosed. A computer system accesses acoustic data for a recorded spoken language and generates a plurality of potential transcriptions for the acoustic data. The computer system scores the plurality of potential transcriptions to create an initial likelihood score for the plurality of potential transcriptions. For a particular potential transcription in the plurality of transcriptions, the computer system generates a syntactical likelihood score. The computer system creates an adjusted score for the particular potential transcription by combining the initial likelihood score and the syntactic likelihood score for the particular potential transcription.

Abstract: Technologies for detecting an end of a sentence in automatic speech recognition are disclosed. An automatic speech recognition device may acquire speech data, and identify phonemes and words of the speech data. The automatic speech recognition device may perform a syntactic parse based on the recognized words, and determine an end of a sentence based on the syntactic parse. For example, if the syntactic parse indicates that a certain set of consecutive recognized words form a syntactically complete and correct sentence, the automatic speech recognition device may determine that there is an end of a sentence at the end of that set of words.

Abstract: Generally, this disclosure provides systems, devices, methods and computer readable media for adaptation of language models and semantic tracking to improve automatic speech recognition (ASR). A system for recognizing phrases of speech from a conversation may include an ASR circuit configured to transcribe a user's speech to a first estimated text sequence, based on a generalized language model. The system may also include a language model matching circuit configured to analyze the first estimated text sequence to determine a context and to select a personalized language model (PLM), from a plurality of PLMs, based on that context. The ASR circuit may further be configured to re-transcribe the speech based on the selected PLM to generate a lattice of paths of estimated text sequences, wherein each of the paths of estimated text sequences comprise one or more words and an acoustic score associated with each of the words.

Abstract: Technologies for detecting an end of a sentence in automatic speech recognition are disclosed. An automatic speech recognition device may acquire speech data, and identify phonemes and words of the speech data. The automatic speech recognition device may perform a syntactic parse based on the recognized words, and determine an end of a sentence based on the syntactic parse. For example, if the syntactic parse indicates that a certain set of consecutive recognized words form a syntactically complete and correct sentence, the automatic speech recognition device may determine that there is an end of a sentence at the end of that set of words.

Abstract: A system and method for real-time processing a signal of a voice interaction. In an embodiment, a digital representation of a portion of an interaction may be analyzed in real-time and a segment may be selected. The segment may be associated with a source based on a model of the source. The model may updated based on the segment. The updated model is used to associate subsequent segments with the source. Other embodiments are described and claimed.

Abstract: System and techniques for direct motion sensor input to rendering pipeline are described herein. A ranked list of ASR hypotheses may be obtained. A set of ASR hypotheses may be selected from the list. The set of ASR hypothesis may be re-ranked using semantic coherence scoring between words in the ASR hypotheses. An ASR hypothesis from the set of ASR hypotheses with a highest re-rank may be outputted.

Abstract: Technologies for detecting an end of a sentence in automatic speech recognition are disclosed. An automatic speech recognition device may acquire speech data, and identify phonemes and words of the speech data. The automatic speech recognition device may perform a syntactic parse based on the recognized words, and determine an end of a sentence based on the syntactic parse. For example, if the syntactic parse indicates that a certain set of consecutive recognized words form a syntactically complete and correct sentence, the automatic speech recognition device may determine that there is an end of a sentence at the end of that set of words.

Abstract: Generally, this disclosure provides systems, devices, methods and computer readable media for adaptation of language models and semantic tracking to improve automatic speech recognition (ASR). A system for recognizing phrases of speech from a conversation may include an ASR circuit configured to transcribe a user's speech to a first estimated text sequence, based on a generalized language model. The system may also include a language model matching circuit configured to analyze the first estimated text sequence to determine a context and to select a personalized language model (PLM), from a plurality of PLMs, based on that context. The ASR circuit may further be configured to re-transcribe the speech based on the selected PLM to generate a lattice of paths of estimated text sequences, wherein each of the paths of estimated text sequences comprise one or more words and an acoustic score associated with each of the words.

Abstract: Various systems and methods for providing a repositionable video display on a mobile device, to emulate the effect of user-controlled binoculars, are described herein. In one example, one or more high resolution video sources (such as UltraHD video cameras) obtain video that is wirelessly broadcasted to mobile devices. The mobile device processes the broadcast based on the approximate location of the spectator's mobile device, relative to a scene within the field of view of the mobile device. The location of the mobile device may be derived from a combination of network monitoring, camera inputs, object recognition, and the like. Accordingly, the spectator can obtain a virtual magnification of a scene from an external video source displayed on the spectator's mobile device.

Abstract: Various systems and methods for providing speech recognition services are described herein. A user device for providing speech recognition services includes a speech module to maintain a speech recognition model of a user of the user device; a user interaction module to detect an initiation of an interaction between the user and a target device; and a transmission module to transmit the speech recognition model to the target device, the target device to use the speech recognition model to enhance a speech recognition process executed by the target device during the interaction between the user and the target device.

Abstract: The subject matter discloses a method for trade interaction chain reconstruction comprising: identifying a swap deal, the swap deal includes two or more of the received interactions and involves two or more participants; selecting a first interaction of the received interactions, said first interaction involves at least two participants of the two or more participants, said first interaction is stored on a computerized device; obtaining a first plurality of interactions of the received interactions that involve the at least two participants of the two or more participants; determining a first plurality of relevance scores between the first plurality of interactions and the first interaction; and associating interactions of the first plurality of interactions to be relevant to the swap deal according to the determined first plurality of relevance scores.

Abstract: A system and method for receiving an initial search entry to search text data. The text data may be, for example, an N word lattice, transcribed by a text to speech engine. The difference between the initial search entry and one or more entries in the dictionary may be measured. One or more similar entries may be selected from the dictionary that have the smallest measures of difference to the initial search entry. The text data may be searched for the one or more selected similar entries. Each of the searched similar entries found in the text data may be displayed as a search result.

Abstract: One embodiment provides an apparatus. The apparatus includes a processor; at least one peripheral device coupled to the processor; a memory coupled to the processor; a generic sentiment model and a first domain training corpus stored in memory; and a hybrid sentiment analyzer logic stored in memory and to execute on the processor. The hybrid sentiment analyzer logic includes a sentiment lexicon generator logic to generate a domain sentiment lexicon based, at least in part, on the first domain training corpus and to store the domain sentiment lexicon in memory, a lexicon-based sentiment classifier logic to generate an annotated training corpus unsupervisedly, based, at least in part, on the domain sentiment lexicon and to store the annotated training corpus in memory, and a model-based sentiment adaptor logic to adapt the generic sentiment model based, at least in part, on the annotated training corpus to generate an adapted sentiment model and to store the adapted sentiment model in memory.