Abstract: A system and method may be configured to analyze audio information derived from an audio signal. The system and method may track sound pitch across the audio signal. The tracking of pitch across the audio signal may take into account change in pitch by determining at individual time sample windows in the signal duration an estimated pitch and a representation of harmonic envelope at the estimated pitch. The estimated pitch and the representation of harmonic envelope may then be implemented to determine an estimated pitch for another time sample window in the signal duration with an enhanced accuracy and/or precision.

Abstract: A method for coding and decoding an audio signal or speech signal and an apparatus adopting the method are provided.

Abstract: An automatic speech recognition system and method are provided for written-domain language modeling.

Abstract: A method for phonetizing text-containing data records that include graphemes includes: phonetizing the data records by converting the graphemes in the data records into phonemes, and storing the phonemes as phonetized data records; and preprocessing to condition the graphemes for the phonetization by modifying the graphemes on a language-defined and/or user-defined basis. The preprocessing of the graphemes and the conversion of the graphemes into phonemes are performed in parallel on different computation units or different portions of the computation units.

Abstract: Disclosed is a device having an audio interface configured to generate from the audio signal an outgoing audio signal for supplying to a loudspeaker component. The audio interface is configured, in generating the outgoing audio signal, to apply dynamic range compression to the audio signal. Device software is configured to receive an incoming audio signal and generate an audio signal from the incoming audio signal. The audio signal generated by the software is supplied to the audio interface for outputting by the loudspeaker component and is also used as a reference in audio signal processing. Generating the audio signal comprises the software applying initial nonlinear amplitude processing to the incoming audio signal to modify its power envelope. The modified power envelope is sufficiently smooth to be substantially unaffected by the dynamic range compression when applied by the audio interface.

Abstract: A semantic tagging method may add context to a sentence in order to increase search efficiency. Regardless of an author's writing style, translating semantic concepts into tags may increase search efficiency. Automatic semantic tagging of documents may allow semantic search and reasoning. Text for semantic tagging may include an email, a website chat room, an internet forum, or a text message. Additional texts may include aggregating general consensus of an emailed topic across multiple emails, whether in the same email chain or separate emails. To increase search efficiency, the analysis of prior communications within the body of text may comprise analyzing structured contextual information to facilitate with homophora resolution. The structured contextual information may include at least one of a sender email address, one or more recipient email addresses, a subject field, a message date and time stamp, and an attachment title.

Abstract: In accordance with an embodiment of the present invention, a method of speech processing included receiving a coded audio signal having coding noise. The method further includes generating a decoded audio signal from the coded audio signal, and determining a pitch corresponding to the fundamental frequency of the audio signal. The method also includes determining the minimum allowable pitch and determining if the pitch of the audio signal is less than the minimum allowable pitch. If the pitch of the audio signal is less than the minimum allowable pitch, applying an adaptive high pass filter on the decoded audio signal to lower the coding noise at frequencies below the fundamental frequency.

Abstract: Disclosed herein are a system, non-transitory computer-readable medium, and method for responding to natural language queries. Keywords likely to appear in a natural language query are determined and each likely keyword is associated with a module. A response to a natural language query comprises information generated by each module associated with a likely keyword appearing in the natural language query.

Abstract: A method for phonetizing a data list having text-containing list entries, each list entry in the data list being subdivided into at least two data fields for provision to a voice-controlled user interface, includes: converting a list entry from a text representation into phonetics; storing the phonetics as phonemes in a phonetized data list; inserting a separating character into the text of a list entry between the respective data fields of the list entry, concomitantly converting the inserted separating character into phonetics and concomitantly storing the converted separating character as a phoneme symbol; and storing the phonemes in a phonetic database, the phonetized data list being produced from the phonemes stored in the phonetic database.

Abstract: Systems and methods for acquiring information associated with a user by using NLP techniques are disclosed. One or more phrases are classified in one or more categories at least partly on the basis of a period for which a product has been used by the user, the user's experience with the product, preferences of the user, or needs of the user by applying one or more natural language processing (NLP) techniques. The one or more phrases are extractable from an electronic publication at least partly on the basis of on a predefined set of verbs, a predefined set of domain-specific terms, and terms indicative of temporal information. One or more terms from the classified phrases are extracted, in which the one or more terms are indicative of the information about the user.

Abstract: Features are disclosed for spotting keywords in utterance audio data without requiring the entire utterance to first be processed. Likelihoods that a portion of the utterance audio data corresponds to the keyword may be compared to likelihoods that the portion corresponds to background audio (e.g., general speech and/or non-speech sounds). The difference in the likelihoods may be determined, and keyword may be triggered when the difference exceeds a threshold, or shortly thereafter. Traceback information and other data may be stored during the process so that a second speech processing pass may be performed. For efficient management of system memory, traceback information may only be stored for those frames that may encompass a keyword; the traceback information for older frames may be overwritten by traceback information for newer frames.

Abstract: Reliable speaker-based clustering of speech utterances allows improved speaker recognition and speaker-based speech segmentation. According to at least one example embodiment, an iterative bottom-up speaker-based clustering approach employs voiceprints of speech utterances, such as i-vectors. At each iteration, a clustering confidence score in terms of Silhouette Width Criterion (SWC) values is evaluated, and a pair of nearest clusters is merged into a single cluster. The pair of nearest clusters merged is determined based on a similarity score indicative of similarity between voiceprints associated with different clusters. A final clustering pattern is then determined as a set of clusters associated with an iteration corresponding to the highest clustering confidence score evaluated. The SWC used may further be a modified SWC enabling detection of an early stop of the iterative approach.

Abstract: The present disclosure provides an information processing method for addressing the technical problem that an operation mode of a voice input interface of a conventional electronic device is not flexible. The method comprises steps of: obtaining a voice input trigger operation by an sensing unit; in response to the voice input trigger operation, starting a voice processing system, and displaying a voice indicator in a display unit, the voice indicator occupying a portion of the display area; obtaining an input operation for moving the voice indicator by the sensing unit; making a response to the input operation for moving; controlling movement of the voice indicator within the display area based on the input operation for moving; determining a corresponding control command based on parameter information of the input operation for moving, the control command being used for controlling a processing procedure of the voice processing system.

Abstract: Techniques disclosed herein include systems and methods for voice-enabled searching. Techniques include a co-occurrence based approach to improve accuracy of the 1-best hypothesis for non-phrase voice queries, as well as for phrased voice queries. A co-occurrence model is used in addition to a statistical natural language model and acoustic model to recognize spoken queries, such as spoken queries for searching a search engine. Given an utterance and an associated list of automated speech recognition n-best hypotheses, the system rescores the different hypotheses using co-occurrence information. For each hypothesis, the system estimates a frequency of co-occurrence within web documents. Combined scores from a speech recognizer and a co-occurrence engine can be combined to select a best hypothesis with a lower word error rate.

Abstract: A wireless telephone includes an adaptive noise canceling (ANC) circuit that adaptively generates an anti-noise signal from a reference microphone signal and injects the anti-noise signal into the speaker or other transducer output to cause cancellation of ambient audio sounds. An error microphone is also provided proximate the speaker to provide an error signal indicative of the effectiveness of the noise cancellation. A secondary path estimating adaptive filter is used to estimate the electro-acoustical path from the noise canceling circuit through the transducer so that source audio can be removed from the error signal. Sidetone is injected into the transducer output, but is not provided to the coefficient control of the secondary path estimating adaptive filter, so that the ambient noise present in the near-end speech microphone signal, and thus present in the sidetone information, does not destabilize the ANC circuit or otherwise cause improper generation of the anti-noise signal.

Abstract: The subject matter discloses a method two phase phonetic indexing and search comprising: receiving a digital representation of an audio signal; producing a phonetic index of the audio signal; producing phonetic N-gram sequence from the phonetic index by segmenting the phonetic index into a plurality of phonetic N-grams; and producing an inverted index of the plurality of phonetic N-grams.

Abstract: Query history expansion may be provided. Upon receiving a spoken query from a user, an adapted language model may be applied to convert the spoken query to text. The adapted language model may comprise a plurality of queries interpolated from the user's previous queries and queries associated with other users. The spoken query may be executed and the results of the spoken query may be provided to the user.

Abstract: A semantic tagging method may add context to a sentence in order to increase search efficiency. Regardless of an author's writing style, translating semantic concepts into tags may increase search efficiency. Automatic semantic tagging of documents may allow semantic search and reasoning. Text for semantic tagging may include an email, a website chat room, an internet forum, or a text message. Additional texts may include aggregating general consensus of an emailed topic across multiple emails, whether in the same email chain or separate emails. To increase search efficiency, the analysis of prior communications within the body of text may comprise analyzing structured contextual information to facilitate with homophora resolution. The structured contextual information may include at least one of a sender email address, one or more recipient email addresses, a subject field, a message date and time stamp, and an attachment title.

Abstract: In particular embodiments, one or more computer-readable non-transitory storage media embody software that is operable when executed to receive an audio waveform fingerprint and a client-determined location from a client device. The received audio waveform fingerprint may be compared to a database of stored audio waveform fingerprints, each stored audio waveform fingerprint associated with an object in an object database. One or more matching audio waveform fingerprints may be found from a comparison set of audio waveform fingerprints obtained from the audio waveform fingerprint database. Location information associated with a location of the client device may be determined, and the location information may be sent to the client device. The client device may be operable to update the client-determined location based at least in part on the location information.

Abstract: A method and an apparatus for encoding and decoding audio signals using adaptive sinusoidal coding are provided. The audio signal encoding method includes the steps of dividing a synthesized audio signal into a plurality of sub-bands, calculating the energy of each sub-band, selecting a predetermined number of sub-bands having a relatively large amount of energy from the sub-bands, and performing sinusoidal coding with regard to the selected sub-bands. Application of sinusoidal coding based on consideration of the amount of energy of each sub-band of the synthesized signal improves the quality of the synthesized signal more efficiently.