Abstract: An audio receiving system includes logic configured to reduce the accumulation of delays caused by the late arrival of audio packets. This logic is configured to accelerate or decelerate presentation of a resulting audio stream in response to the detection of late packets. The acceleration is discontinued once the effects of the late packets have been compensated for. The audio receiving system is typically applied to applications in which lag is undesirable. These can include web conferencing, telepresence, and online video games.

Abstract: A system and method for adding audio indicators, such as an audio tone, to audio data corresponding to text. The audio indicator corresponds to secondary content in text, such as a footnote. The system will insert and/or modify a textual indicator at a location in the text corresponding to the secondary content. The system will then process the textual indicator during text-to-speech processing of the text so that an audio tone is output at a moment in speech data corresponding to the location of the secondary content in the text. A speech synthesis markup language may be used to create and/or modify the textual indicator.

Abstract: Methods and systems for enrolling a user in an authentication program. In some embodiments, voice interaction that includes a request or command is received from a user. The user may be requested to provide authentication information to fulfill the request or command made during the voice interaction. The user may be authenticated using a first authentication method. The user may be passively enrolled into an authentication program that uses a second authentication method. Enrolling may include deriving characteristics of the user's voice from the voice interaction. After the user is enrolled in the authentication program, the second authentication method may be used to authenticate the user prior to fulfilling requests or commands made during voice navigation.

Abstract: A minutes making assistance device according to the present invention includes: a sound processing unit that performs processing regarding a voice and determines whether or not speaking is started; an operation processing unit that performs processing regarding an operation and determines whether or not the operation is performed; a display processing unit that performs processing regarding a display; and a control unit that stores speaking start time and warning time in a memory when the sound processing unit determines that the speaking is started, performs warning processing when the current time becomes the warning time, and terminates the processing when the operation processing unit determines that the operation is performed before the warning time.

Abstract: Automated, standardized and accurate extraction of relationships within text. Automatic extraction of such relationships/information allows the information to be stored in structured form so that it can be easily and accurately retrieved when needed. Such information can be used to build online search engines for highly specific and accurate information retrieval. Generally, according to the current invention, extracting such information (i.e., relationships within text) from raw text can be accomplished using natural language processing (NLP) and graph theoretic algorithm. Examples of such textual relationships include, but are not limited to, biological relationships between biological terms such as proteins, genes, pathways, diseases and drugs. The current methodology is also able to recognize negative dependences in context, match patterns, and provide a shortest path between related words.

Abstract: Techniques are described herein that are capable of satisfying specified intent(s) based on multimodal request(s). A multimodal request is a request that includes at least one request of a first type and at least one request of a second type that is different from the first type. Example types of request include but are not limited to a speech request, a text command, a tactile command, and a visual command. A determination is made that one or more entities in visual content are selected in accordance with an explicit scoping command from a user. In response, speech understanding functionality is automatically activated, and audio signals are automatically monitored for speech requests from the user to be processed using the speech understanding functionality.

Abstract: Systems and methods for providing non-lexical cues in synthesized speech are described herein. Original text is analyzed to determine characteristics of the text and/or to derive or augment an intent (e.g., an intent code). Non-lexical cue insertion points are determined based on the characteristics of the text and/or the intent. One or more nonlexical cues are inserted at insertion points to generate augmented text. The augmented text is synthesized into speech, including converting the non-lexical cues to speech output.

Abstract: A system, apparatus, and method for verifying that a consumer seeking to conduct a transaction with a merchant is authorized to use an associated account. An exemplary embodiment of the apparatus may include a voice response unit, a storage medium, and a voice print comparator The voice response unit may be configured to obtain a test voice print during a transaction between the consumer and the merchant. The storage medium may be configured to store information associated with the consumer, and a control voice print associated with the information. Also, the voice print comparator may be configured to receive the control voice print, compare the test voice print to the control voice print, and to return a match level signal indicating a degree of match therebetween.

Abstract: A system receives monaural sound which includes speech and background noises. The received sound is divided by frequency and time into time-frequency units (TFUs). Each TFU is classified as speech or non-speech by a processing unit. The processing unit for each frequency range includes at least one of a deep neural network (DNN) or a linear support vector machine (LSVM). The DNN extracts and classifies the features of the TFU and includes a pre-trained stack of Restricted Boltzmann Machines (RBM), and each RBM includes a visible and a hidden layer. The LSVM classifies each TFU based on extracted features from the DNN, including those from the visible layer of the first RBM, and those from the hidden layer of the last RBM in the stack. The LSVM and DNN include training with a plurality of training noises. Each TFU classified as speech is output.

Abstract: An image processing apparatus, which includes a voice input receiver configured to receive a voice input of user, a signal processor configured to recognize and process the received voice input received through the voice input receiver, a buffer configured to store the voice input, and a controller configured to determine whether a voice recognition function of the signal processor is activated and control the signal processor to recognize the voice input stored in the buffer in response to the voice recognition function being determined to be activated. The controller is further configured to store the received voice input in the buffer in response to the received voice input being input through the voice input receiver while the voice recognition function is not activated, so that the received voice input is recognized by the signal processor when the voice recognition function is activated.

Abstract: Provided are a method and apparatus for encoding and decoding a stereo signal or a multi-channel signal. According to the method and apparatus, a stereo signal or a multi-channel signal can be encoded and/or decoded by generating parameters based on a mono signal.

Abstract: According to one embodiment, a speech translation system includes a first terminal device including a first speech input unit for inputting a first speech of a first language spoken by a first user, and converting the first speech to a first speech signal; a second terminal device including a second speech input unit for inputting a second speech of a second language spoken by a second user, and converting the second speech to a second speech signal; a speech recognition device that receives the first speech signal and the second speech signal, recognizes the first speech signal to a first recognized text, and recognizes the second speech signal to a second recognized text; a machine translation device that receives the first recognized text and the second recognized text, translates the first recognized text to a first translated text of the second language, and translates the second recognized text to a second translated text of the first language; and a control device; wherein the first terminal device rec

Abstract: A dialog system uses an extended domain in order to have a dialog with a user using natural language. If a dialog pattern actually input by the user is different from a dialog pattern predicted by an expert, an extended domain generated in real time based on user input is used and an extended domain generated in advance is used to have a dialog with the user.

Abstract: A construction and display of speech commands system that allows a user to simply read what is on an application that involves visual elements with which the user interacts, and in doing so, gives the appropriate commands to the speech recognition system for the task at hand. The construction and display of speech commands system may include a speech recognition system, a grammar builder module, and a speech enablement module. The construction and display of speech commands system may automatically generate a speech enabled application from generated speech grammar.

Abstract: A method for verifying and identifying users, and for verifying users' identity, by means of an authentication device capable of transmitting, receiving and recording audio or ultrasonic signals, and capable of converting the signals into digital data, and performing digital signal processing. Voice pattern(s) and user(s) information of one or more authorized user(s) are recorded and stored on the authentication device. User(s) identification is verified by inputting to the authentication device a vocal identification signal from a user, and comparing the voice pattern of the vocal identification signal with the recorded voice pattern(s) of the authorized user(s), and if a match is detected issuing an indication that the user is identified as an authorized user.

Abstract: A speech server (SS) managing one or a plurality of pieces of speech terminal-specifying information (STSI) and user-specifying information (USI), each of pieces of STSI allowing a corresponding one of one or a plurality of speech terminals to be specified, USI being of a user who is capable of causing the corresponding one of the one or a plurality of speech terminals to output speech. The SS receives USI and transmit the one or a plurality of pieces of STSI associated with USI. The SS receives (i) STSI selected from the one or a plurality of pieces of STSI transmitted and (ii) speech information indicative of speech content to be outputted as speech. The SS instructs a speech terminal to output the speech content as speech, the speech terminal being identified among the one or a plurality of speech terminals by STSI received.

Abstract: A method for detecting data in a sequence of characters or text using both a statistical engine and a pattern engine. The statistical engine is trained to recognize certain types of data and the pattern engine is programmed to recognize the grammatical pattern of certain types of data. The statistical engine may scan the sequence of characters to output first data, and the pattern engine may break down the first data into subsets of data. Alternatively, the statistical engine may output items that have a predetermined probability or greater of being a certain type of data and the pattern engine may then detect the data from the output items and/or remove incorrect information from the output items.

Abstract: A communication system with a base station configured to determine a codec to use with end units, such that, in response to a determination that a first end unit uses a first set of access information, the base station registers the first end unit to the base station, setting the first codec to be used for communications with the first end unit, and, in response to a determination that a second unit uses a second set of access information, the base station registers the second end unit to the base station, setting the second codec to be used for communications with the second end unit. The communication system also comprises an end unit configured to determine the codec used by the base station and set the determined codec as the codec to use for communications with the base station.

Abstract: A speech processing apparatus that collects sound signals. With each of the collected sound signals, the apparatus may estimate a direction of a sound source, and select an extension filter that is applied to each sound signal. The extension filter may correspond to the estimated sound source of each of the sound signals. In addition, each of the sound signals may be corrected using the extension filter, and a reverberation reduction of the corrected sound signals and the collected sound signals may be performed.

Abstract: A low power sound recognition sensor is configured to receive an analog signal that may contain a signature sound. Sparse sound parameter information is extracted from the analog signal and compared to a sound parameter reference stored locally with the sound recognition sensor to detect when the signature sound is received in the analog signal. A portion of the sparse sound parameter information is differential zero crossing (ZC) counts. Differential ZC rate may be determined by measuring a number of times the analog signal crosses a threshold value during each of a sequence of time frames to form a sequence of ZC counts and taking a difference between selected pairs of ZC counts to form a sequence of differential ZC counts.