Abstract: A pick up irradiates light beams at specified intervals in the tangential direction of the main track to be reproduced and both of adjacent tracks on a disk 10 on which tracks are formed. A CTC unit 15 uses the delay amount ?d of the respective sample-value series corresponding to the reproduction signal RFm from the main track and reproduction signals RF1, RF2 from both adjacent tracks to correct the delay, then outputs a CTC output signal from which the cross-talk component has been removed. In addition, when adjusting the delay, the CPU 17 applies a disturbance to the control signal for the actuator of the servo-control unit 18 and changes the delay amount ?d within a specified range of change. At this time, the CPU 17 sets the delay amount ?d, which minimizes the jitter value found by the jitter detection unit 16 according to the CTC output signal, for the CTC unit 15. This makes it possible to set the optimum delay amount ?d when adjusting the delay, even when there is little steady cross-talk.

Abstract: An optical pickup unit for scanning an optical disk includes a first optical branch including a first radiation source for emitting a first radiation beam of a first wavelength, and a dichroic mirror (8) located in the path of the first radiation beam for reflecting the first beam towards the disk, and a second optical branch including a second radiation source for emitting a second radiation beam of a second wavelength, different to said first wavelength, and a folding mirror (28) located in the path of the second radiation beam for reflecting the second beam towards the disk. The first and second branches are stacked in the axial direction of the disk, and are arranged substantially perpendicularly when viewed along the axial direction.

Abstract: A system and method for detecting beats in a compressed audio domain is disclosed where a beat detector functions as part of an error concealment system in an audio decoding section used in audio information transfer and audio download-streaming system terminal devices such as mobile phones. The beat detector includes a MDCT coefficient extractor, a band feature value analyzer, a confidence score calculator; and a converging and storage unit. The method provides beat detection by means of beat information obtained using both MDCT coefficients as well as window-switching information. A baseline beat position is determined using MDCT coefficients obtained from the audio bitstream which also provides a window-switching pattern. A window-switching beat position is compared with the baseline beat position and, if a predetermined condition is satisfied, the window-switching beat position is validated as a detected beat.

Abstract: An apparatus detects the number of tracks traversed on an optical disk. The apparatus includes a differentiation part for differentiating an output signal of a bandpass filter provided on an adjacent stage which eliminates a low DC component, a high frequency component and noise of an inputted modulation signal. A sample trigger generating circuit generates top and bottom signals in response to a data signal outputted through the differentiation part and a signal outputted through a comparator. A top hold part and a bottom hold part holds the data signal at top and bottom. A subtraction part performs a subtraction to obtain a difference between the top and bottom signals outputted from the top and bottom hold parts, thereby resulting in counting the number of the tracks traversed, while increasing a jumping speed required for a detection of the number of the tracks traversed.

Abstract: System, method and computer-readable medium containing instructions for providing AV signals with open or closed captioning information. The system includes a speech-to-text processing system coupled to a signal separation processor and a signal combination processor for providing automated captioning for video broadcasts contained in AV signals. The method includes separating an audio signal from an AV signal, converting the audio signal to text data, encoding the original AV signal with the converted text data to produce a captioned AV signal and recording and displaying the captioned AV signal. The system may be mobile and portable and may be used in a classroom environment for producing recorded captioned lectures and used for broadcasting live, captioned lectures. Further, the system may automatically translate spoken words in a first language into words in a second language and include the translated words in the captioning information.

Abstract: Media encoding, transmission, and playback processes and structures employ a multi-channel architecture with different audio channels corresponding to different playback rates for a presentation to be transmitted over a network. Audio frames in the various audio channels all correspond to the same amount of time in the original presentation and have frame indexes that identify in the different audio channels the frames corresponding to the same time interval in the presentation. A user can make a real-time change in playback rate causing selection of a channel corresponding to the new playback rate and a frame required for prompt and smooth transition in the playback rate of the presentation. The architecture can additionally provide channels for graphics data such as image data that are displayed according to the index of the audio, and different audio channels with the same playback rate but different compression schemes for use according to available bandwidth on the network.

Abstract: As a retrieval result, appropriate text of a second language is provided in response to a retrieval request by text of a first language. A first directory storing part stores a first directory structure created for a first language. A second directory storing part stores a second directory structure created for a second language. A directory relation storing part stores correspondences between directories in the first directory structure and directories in the second directory structure. A directory retrieval part receives a retrieval request by the first language from a user and decides which directory in the first directory structure the request has a high degree of relation with. A multilingual retrieval part decides documents having a high degree of relation with the retrieval request, of documents belonging to a directory in the second directory structure that corresponds to the decided directory.

Abstract: A method of initializing an ITU Recommendation G.729 Annex B compliant voice activity detection (VAD) device is disclosed, having the steps of (1) determining a first set of running average background noise characteristics in accordance with Recommendation G.729B; (2) determining a second set of running average background noise characteristics; and (3) substituting the second set of running average background noise characteristics for the first set when a specific event occurs. The specific event is a divergence between the first and second sets of running average background noise characteristics.

Abstract: Methods and systems which provide for standardized inpatient-outpatient nomenclatures and accepting both outpatient and inpatient data to commonly accessible storage. In a first method embodiment, the first method is characterized by accepting user input identifying at least two different names for a substantially similar grouping of one or more medical criteria; and accepting user input specifying at least one of the at least two different names as forming at least a part of an outpatient-inpatient standardized nomenclature for the substantially similar grouping of one or more medical criteria. In a second method embodiment, the second method is characterized by accepting either outpatient or inpatient data to commonly accessible storage.

Abstract: In order to improve recognition performance, a no-speech sound model correction section performs an adaptation of a no-speech sound model which is a sound model representing a no-speech state on the basis of input data observed in an interval immediately before a speech recognition interval for the object of speech recognition and the degree of freshness representing the recentness of the input data.

Abstract: In a text-to-speech system, a method of converting text-to-speech can include receiving a text input and comparing the received text input to at least one entry in a text-to-speech cache memory. Each entry in the text-to-speech cache memory can specify a corresponding spoken output. If the text input matches one of the entries in the text-to-speech cache memory, the cached speech output specified by the matching entry can be provided.

Abstract: Speech recognition methods, systems, and interfaces are used in the generation of medical reports from data in a hierarchically-organized database for the entry and searching of data in a database based on spoken utterances of a user. A workflow function facilitates a series of contexts, typically based on information in a knowledge base, that are used to establish procedural rules and word-mapping databases for each context for word-matching data entry based on spoken utterances of a user. The generation of medical reports from the entered medical data provides for searching the database generated using the speech recognition methods. The series of contexts and word-mapping databases are developed using a hierarchically-organized database representation based on knowledge regarding the relationship of data items in the main database.

Abstract: A Monte Carlo method for use with natural language understanding and speech recognition language models can include a series of steps. The steps can include identifying at least one phrase embedded in a body of text wherein the phrase can belong to a phrase class. An additional attribute corresponding to the identified phrase can be determined. The body of text can be copied and the identified phrase can be replaced with a different phrase selected from a plurality of phrases. The different phrase can belong to the phrase class and correspond to the attribute.

Abstract: A speech encoding/decoding method using an encoder working at very low bit rates, comprises a learning step enabling the identification of the representatives of the speech signal; and an encoding step to segment the speech signal and determine the best representative associated with each recognized segment. The method also comprises at least one step for the encoding/decoding of at least one of the parameters of the prosody of the recognized segments, e.g., the energy, pitch, voicing, and/or length of the segments, by using a piece of information on prosody pertaining to the best representatives. The method can employ a bit rate of lower than 400 bits per second.

Abstract: A system and method is provided that randomly generates text with a given structure. The structure is taken from a number of learning examples. The structure of training examples is captured by word classification and the definition of the relationships between word classes in a given language. The text generated with this procedure is intended to replicate the information given by the original learning examples. The resulting text may be used to better model the structure of a language in a stochastic language model.

Abstract: A method and apparatus are provided for compressing and using a concatenative speech database in TTS systems to improve the quality of speech output generated by handheld TTS systems by allowing synthesis to occur on the client. According to one embodiment of the present invention, a G.723 encoder receives diphone waveforms, and compresses them into diphone residuals. While compressing the diphone waveforms, the encoder generates Linear Predictive Coding (LPC) coefficients. The diphone residuals, and the encoder-generated LPC coefficients are then stored in encoder-generated compressed packet.

Abstract: A method for speech synthesis includes receiving an input speech signal containing a set of speech segments, and estimating spectral envelopes of the input speech signal in a succession of time intervals during each of the speech segments. The spectral envelopes are integrated over a plurality of window functions in a frequency domain so as to determine elements of feature vectors corresponding to the speech segments. An output speech signal is reconstructed by concatenating the feature vectors corresponding to a sequence of the speech segments.

Abstract: A system and method of providing sender customization of multi-media messages through the use of inserted images or video. The images or video may be sender-created or predefined and available to the sender via a web server. The method relates to customizing a multi-media message created by a sender for a recipient, the multi-media message having an animated entity audibly presenting speech converted from text created by the sender. The method comprises receiving at least one image from the sender, associating each at least one image with a tag, presenting the sender with options to insert the tag associated with one of the at least one image into the sender text, and after the sender inserts the tag associated with one of the at least one images into the sender text, delivering the multi-media message with the at least one image presented as background to the animated entity according to a position of the tag associated with the at least one image in the sender text.

Abstract: A trained vector generation section 16 generates beforehand a trained vector v of unvoiced sounds. An LPC Cepstrum analysis section 18 generates a feature vector A of a voice within the non-voice period, an inner product operation section 19 calculates an inner product value VTA between the feature vector A and the trained vector V, and a threshold generation section 20 generates a threshold ?v on the basis of the inner product value VTA. Also, the LFC Cepstrum analysis section 18 generates a prediction residual power ? of the signal within the non-voice period, and the threshold generation section 22 generates a threshold THD on the basis of the prediction residual power ?.

Abstract: The present invention relates to a speech recognition apparatus and a speech recognition method for speech recognition with improved accuracy. A distance calculator 47 determines the distance from a microphone 21 to a user uttering. Data indicating the determined distance is supplied to a speech recognition unit 41B. The speech recognition unit 41B has plural sets of acoustic models produced from speech data obtained by capturing speeches uttered at various distances. From those sets of acoustic models, the speech recognition unit 41B selects a set of acoustic models produced from speech data uttered at a distance closest to the distance determined by the distance calculator 47, and the speech recognition unit 41B performs speech recognition using the selected set of acoustic models.