Abstract: A speech recognition system which effectively recognizes unknown speech from multiple acoustic environments includes a set of secondary models, each associated with one or more particular acoustic environments, integrated with a base set of recognition models. The speech recognition system is trained by making a set of secondary models in a first stage of training, and integrating the set of secondary models with a base set of recognition models in a second stage of training.

Abstract: A pattern adapting apparatus including an input pattern forming unit, a tree structure standard pattern storing unit for storing a tree structure standard pattern including a tree structure indicative of inclusive relationships among categories and a parameter set at each node of the tree structure, a pattern matching unit for matching categories of the tree structure standard pattern with input samples of an input pattern, a tree structure standard pattern modifying unit for modifying a tree structure standard pattern based on the results of pattern matching, a node set selecting unit for calculating a description length with respect to a plurality of node sets in a tree structure pattern to select an appropriate node set, a modified standard pattern forming unit for forming a modified standard pattern by using a parameter set of a selected node set, and a standard pattern for recognition storing unit for storing a modified standard pattern.

Abstract: Foe executing the code excitation linear prediction (CELP) coding, for example, .alpha.-parameters are taken out from the input speech signal by a linear prediction coding (LPC) analysis circuit 12. The .alpha.-parameters are then converted by an .alpha.-parameter to LSP converting circuit 13 into linear spectral pair (LSP) parameters and a vector of these line spectral pair (LSP) parameters is vector-quantized by a quantizer 14. The changeover switch 16 is controlled depending upon the pitch value detected by a pitch detection circuit 22 for selecting and using one of the codebook 15M for male voice and the codebook 15F for female voice for improving quantization characteristics without increasing the transmission bit rate.

Abstract: The present invention relates to a word-spotting system and a method for finding a keyword in ascoustic data. The method includes a filler recognition phase and a keyword recognition phase wherein: during the filler recognition phase the acoustic data is processed to identify phones and to generate temporal delimiters and likelihood scores for the phones; during the keyword recognition phase, the acoustic data is processed to identify instances of a specified keyword including a sequence of phones; wherein the temporal delimiters and likelihood scores generated in the filler recognition phase are used in the keyword recognition phase.

Abstract: In a transmission system a signal is encoded in an encoder (7) and the encoded signal is transmitted by a transmitter (2) via a transmission medium (4) to a receiver (6). In the encoder, analysis parameters of the input signal are determined by an analyzer (8) and quantized by a quantizer (14). The transmitter transmits quantization level numbers to the receiver (6), and in the receiver decoded analysis parameters are derived by interpolating level numbers of two subsequent sets of analysis parameters, and subsequently determining the corresponding analysis. By interpolating the level numbers instead of the analysis parameters themselves, a substantial amount of computational complexity is saved.

Abstract: A voice recognition interactive conversational apparatus employs temporally related data storage tracks containing interactive conversational content. The conversational content is stored on each track in information segments. Each of the segments forms a complete message reproducible in response to the selection of the track upon which the segments are stored. Each information segment includes interrogatories having vocal responses, responsive messages, informational messages, or combinations thereof related in time and content to information contained on at least one other track. The information is stored on the tracks in a predetermined timed sequence for providing interactive conversation dependent on the verbal responses. A playback device facilitates playback of the conversational content. A track selector randomly accesses the tracks for retrieving and play back the information stored thereon via the playback device.

Abstract: Frequency warping approaches to speaker normalization have been proposed and evaluated on various speech recognition tasks. In all cases, frequency warping was found to significantly improve recognition performance by reducing the mismatch between test utterances presented to the recognizer and the speaker independent HMM model. This invention relates to a procedure which compensates utterances by simultaneously scaling the frequency axis and reshaping the spectral energy contour. This procedure is shown to reduce the error rate in a telephone based connected digit recognition task by 30%.

Abstract: A personal reach system includes a network, a personal reach unit, a cellular network, and a landline telephone station. The personal reach unit is coupled to the cellular network and the landline telephone through the network. When a first party pages a second party through the personal research unit, the first party is connected to the second party by either the personal reach unit through the landline telephone station or the cellular network through the cellular phone.

Abstract: A data terminal is able to use data services of a cellular network so that it is detachably connected to a mobile phone with a connection cable. The data services need not be supported by the telephone. The connection cable forms a fast synchronous serial bus. The service-specific data protocols needed in implementing a data service are accomplished in a microprocessor unit in an external interface adaptor connected to one end of the serial bus, said processor being via the bus directly connected to the processor performing the signalling and control functions and the channel coder of the phone. Thus, the interface adaptor is able to exchange messages directly with the processor unit and the channel coder of the phone.

Abstract: A speech packet transmission system of the present invention has a speech coding device and a speech decoding device. In the coding device a speech coder (21) codes a PCM (Pulse Code Modulation) speech and outputs the resulting speech coded data and prediction coefficients. A speech detector (22) determines whether the input PCM speech is voiced or unvoiced. A transmission prediction coefficient memory (23) memorizes the prediction coefficients. A delay circuit (24) delays the speech coded data by a preselected delay time. On the transition from an unvoiced state to a voiced state, a transmitter (25) sends the prediction coefficients as a leading packet and then sends the delayed speech coded data as the following packet. In the decoding device, a receiver (31) received the packets from the coding device separates the packets into the prediction coefficients and the coded data.

Abstract: The present invention provides a method and apparatus for selectively restricting only forwarded calls. Upon realizing that an incoming call was incorrectly forwarded, the terminal receiving the incoming call transmits to a telecommunications exchange an indication requesting activation of a subscriber feature that will restrict only forwarded calls from the terminal forwarding the call. The directory number for the terminal forwarding the call is retrieved from a register. That retrieved directory number is correlated with the directory number representing the terminal incorrectly receiving the forwarded call and the correlated data is stored at a database. Upon the receipt of the next incoming call, the telecommunications exchange determines that the call is forwarded. Then, the directory number associated with the forwarding party is compared against the data previously correlated and stored in the database.

Abstract: The present invention provides a method and apparatus that places visual markers onto the display of a document on a graphical user interface. The visual markers assist in the navigation around the document by numbered artifacts.

Abstract: A time-sequential input pattern (20), which is derived from a continual physical quantity, such as speech is recognized. The system includes input means (30), which accesses the physical quantity and therefrom generates a sequence of input observation vectors. The input observation vectors represent the input pattern. A reference pattern database (40) is used for storing reference patterns, which consist of a sequence of reference units. Each reference unit is represented by associated reference probability densities. A tree builder (60) represents for each reference unit the set of associated reference probability densities as a tree structure. Each leaf node of the tree corresponds to a reference probability density. Each non-leaf node corresponds to a cluster probability density, which is derived from all reference probability densities corresponding to leaf nodes in branches below the non-leaf node.

Abstract: A radio telecommunications network for providing redundant interfaces from a mobile switching center (MSC) to an associated voice mail system (VMS) is disclosed. The radio telecommunications network includes a first peripheral equipment gateway (PEG-1) connected to the VMS, wherein the PEG-1 provides a first interface between the VMS and the radio telecommunications network, and a second peripheral equipment gateway (PEG-2) connected to the VMS, wherein the PEG-2 provides a second interface between the VMS and the radio telecommunications network. The MSC receives and routes an incoming call from a mobile station requesting voice mail retrieval, and a home location register (HLR) simultaneously sends routing request messages to PEG-1 and PEG-2, and includes an indication that the routing request messages are for voice mail retrieval. PEG-1 and PEG-2 return routing numbers for the VMS to the HLR.

Abstract: An apparatus and method, in a telecommunications network having a first and a second system, for activating and deactivating call forwarding from a dual mode subscriber station's number in the second system to the subscriber station's number in the first system. A first message is transmitted upon initiation of communication between the subscriber station and the first system. If it is determined that a data field in the first message is in the first state, call forwarding from the subscriber number of the second system to the subscriber number of the first system is activated. If the data field is in a second state, call forwarding from the subscriber number of the second system to the subscriber number of the first system is not activated. An indication of at least one type of selected subsequent event, upon which the user of the subscriber station desires to deactivate the call forwarding may also be included in the first message. If the selected event occurs, the call forwarding will be deactivated.

Abstract: A speech recognition technique uses lexical tree pre-filtering to obtain lists of words for use in performing speech recognition. The lexical tree pre-filtering includes representing a vocabulary of words using a lexical tree and identifying a first subset of the vocabulary that may correspond to speech spoken beginning at a first time by propagating through the lexical tree information about the speech spoken beginning at the first time. A second subset of the vocabulary that may correspond to speech spoken beginning at a second time is identified by propagating through the lexical tree information about the speech spoken beginning at the second time. Words included in the speech are recognized by comparing speech spoken beginning at the first time with words from the first subset of the vocabulary and speech spoken beginning at the second time with words from the second subset of the vocabulary. The state of the lexical tree is not reset between identifying the first and second subsets.

Abstract: A mobile subscriber enters a service code followed by a called party number. An application module associated with a mobile station stores the received called party number at an attached Subscriber Identity Module (SIM) card. After determining that the mobile station has traveled into the designated home zone, the application module retrieves the stored called party number and automatically originates an outgoing call towards the called party number. By waiting until the mobile station has entered the home zone, the resulting call connection avoids additional air time charges.

Abstract: A noise reducing method and apparatus for reducing quantization error or noise generated on quantizing input signals, such as digital audio signals, wherein a dither signal concentrated in signal energy in the low frequency range is summed to the input signal and the resulting sum signal is sent to a quantizer. The quantizer error generated in the quantizer is fed back via a noise filter to the input side of the quantizer. The resulting signal has a smoothed noise floor and has noise components suppressed in a frequency range to which the human hearing mechanism exhibits high sensitivity, thus realizing the sound quality comparable to that of the input signal which prevailed prior to noise shaping.

Abstract: A signal encoding method and apparatus by which a high encoding efficiency may be achieved by encoding tonal components of plural channels by exploiting channel-to-channel characteristics. The encoding of separately encoding first signals of plural channels is selectively changed over to the encoding of encoding the first signals of plural channels in common or vice versa depending on the results of detection of characteristics of the first signals of plural channels. The compression ratio may be improved if the first signals are encoded in common, while ill effects due to encoding in common may be avoided by separately encoding the first signals of plural channels.