Abstract: A facility is provided for allowing a caller to place a telephone call by merely uttering a label identifying a desired called destination and to charge the telephone call to a particular billing account by merely uttering a label identifying that account. Alternatively, the caller may place the call by dialing or uttering the telephone number of the called destination or by entering a speed dial code associated with that telephone number. The facility includes a speaker verification system which employs cohort normalized scoring. Cohort normalized scoring provides a dynamic threshold for the verification process making the process more robust to variation in training and verification utterences. Such variation may be caused by, e.g., changes in communication channel characteristics or speaker loudness level.

Abstract: A decoding method and apparatus for speech coding systems which takes into account the fact that the human auditory system is sensitive to changes in signal characteristics. For example, a sustained distortion of the spectral characteristic of reconstructed speech is usually less perceptible than an objectively smaller distortion which changes as a function of time. This property of the auditory system is advantageously exploited in the design of a speech coding system receiver in accordance with the present invention, by selecting the sequence of decoded speech parameter values on a perceptual basis. Illustratively, the sequence of decoded speech parameter values is selected so as to describe a smooth path through the sequence of Voronoi regions. The distance between successive parameter values is advantageously minimized, under the constraint that the resultant parameter values fall within, or nearly within, the appropriate Voronoi regions.

Abstract: A speech coding system employing an adaptive codebook model of periodicity is augmented with a pitch-predictive filter (PPF). This PPF has a delay equal to the integer component of the pitch-period and a gain which is adaptive based on a measure of periodicity of the speech signal. In accordance with an embodiment of the present invention, speech processing systems which include a first portion comprising an adaptive codebook and corresponding adaptive codebook amplifier and a second portion comprising a fixed codebook coupled to a pitch filter, are adapted to delay the adaptive codebook gain; determine the pitch filter gain based on the delayed adaptive codebook gain, and amplify samples of a signal in the pitch filter based on said determined pitch filter gain. The adaptive codebook gain is delayed for one subframe. The pitch filter gain equals the delayed. adaptive codebook gain, except when the adaptive codebook gain is either less than 0.2 or greater than 0.8.

Abstract: A low-bitrate (typically 8 kbit/s or less), low-delay digital coder and decoder based on Code Excited Linear Prediction for speech and similar signals features backward adaptive adjustment for codebook gain and short-term synthesis filter parameters and forward adaptive adjustment of long-term (pitch) synthesis filter parameters. A highly efficient, low delay pitch parameter derivation and quantization permits overall delay which is a fraction of prior coding delays for equivalent speech quality at low bitrates.

Abstract: A method identifies and models the distortion in a document using a two-dimensional nonlinear filter. The two-dimensional nonlinear filter is then used to improve the performance of optical character recognition systems in recognizing characters in documents that have been degraded, as for example by repeated photocopy reproductions or by fax transmission.

Abstract: The invention is used to shape noise in time domain and frequency domain coding schemes. The method advantageously uses a noise weighting filter based on a filterbank with variable gains. A method is presented for computing the gains in the noise weighting filterbank with filter parameters derived from the masking properties of speech. Illustrative embodiments of the method in various coding schemes are illustrated.

Abstract: A method and apparatus for recognizing connected and degraded text is presented. A text enhancement process which may be used in combination with text recognition comprises filtering a scanned image to determine whether a binary image value of an image pixel should be complemented, determining whether complementing the value of the pixel reduces the sharpness of wedge-like figures in the image, and complementing the binary value of the pixel when doing so does not reduce sharpness. Image preprocessing may also provide a line adjacency graph (LAG) based on a set of image pixels and the removal of a path from a LAG when the path is located either at the top or bottom of the set of image pixels and is of degree (1,0) or (0,1), respectively, and connected to a junction of the LAG.

Abstract: A method of increasing the testability of sequential circuit designs with use of a clock transformation technique. Circuit states which are difficult to reach, but are nonetheless required to detect at least one fault of the circuit, are automatically identified. In accordance with one illustrative embodiment, estimations of joint line probabilities are compared with a preselected threshold value to identify hard-to-reach states. Then, commonly clocked flip-flops which must be simultaneously assigned values in order to reach the identified states are partitioned into independently clocked groups of flip-flops. In this manner, hard-to-reach circuit states are transformed into easy-to-reach states, which, in turn, results in transforming difficult-to-detect faults into easy-to-detect faults.

Abstract: A speech coding system robust to frame erasure (or packet loss) is described. Illustrative embodiments are directed to a modified version of CCITT standard G.728. In the event of frame erasure, vectors of an excitation signal are synthesized based on previously stored excitation signal vectors generated during non-erased frames. This synthesis differs for voiced and non-voiced speech. During erased frames, linear prediction filter coefficients are synthesized as a weighted extrapolation of a set of linear prediction filter coefficients determined during non-erased frames. The weighting factor is a number less than 1. This weighting accomplishes a bandwidth-expansion of peaks in the frequency response of a linear predictive filter. Computational complexity during erased frames is reduced through the elimination of certain computations needed during non-erased frames only.

Abstract: A method of making a speech recognition model database is disclosed. The database is formed based on a training string utterance signal and a plurality of sets of current speech recognition models. The sets of current speech recognition models may include acoustic models, language models, and other knowledge sources. In accordance with an illustrative embodiment of the invention, a set of confusable string models is generated, each confusable string model comprising speech recognition models from two or more sets of speech recognition models (such as acoustic and language models). A first scoring signal is generated based on the training string utterance signal and a string model for that utterance, wherein the string model for the utterance comprises speech recognition models from two or more sets of speech recognition models. One or more second scoring signals are also generated, wherein a second scoring signal is based on the training string utterance signal and a confusable string model.

Abstract: A method and apparatus for the processing of magnetic recording data signals which provides for the cancellation of the effects of precursor intersymbol interference (ISI) in the data signals. According to one illustrative embodiment, an adaptive filter is used to produce a replica of the precursor ISI, which is then subtracted from a delayed version of the data signal. The amount of the delay is advantageously equal to the delay between the main peak of the channel impulse response and the leading undershoot (i.e., the precursor ISI effect) which is to be canceled. The signal which is provided to the adaptive filter (to produce the precursor ISI replica) may be based on tentative decisions generated by a symbol-by-symbol detector, and may first be converted from three levels to two levels by a three-to-two level converter.

Abstract: An air moving device such as an axial cooling fan comprises a rotatable shaft assembly and a plurality of fan blades coupled to the rotatable shaft assembly, each of the fan blades having a trailing edge when the rotatable shaft assembly is rotated in a given direction. In accordance with one illustrative embodiment of the invention, at least one of the fan blades has a sound reducing material attached thereto alongside the trailing edge thereof. The sound reducing material reduces aeroacoustic noise by reducing the aerodynamic effect that occurs when turbulence produced by the pressure differential between the two sides of one blade comes into contact either with the trailing edge of the given blade which created it, or, more significantly, with the following blade, at or near that blade's trailing edge.

Abstract: A security node disposed in the telecommunications network connecting calling and called parties transforms information (which can be voice, data, facsimile, video and other types of calls or messages) encrypted in a first format to (a) encrypted information in a different format or to (b) non-encrypted information, and vice-versa. The node is accessible from any location connected to the network. By routing calls or messages originated by the calling party and destined for the called party via the security node, and providing appropriate control signals to the node, the information may be encrypted only over a portion of the transmission path between the parties, and clear over the remainder of the transmission path. Alternatively, the information may be encrypted in different portions of the path using different encryption algorithms.

Abstract: A method for making data objects having hidden pointers persistent is disclosed. A pre-existing process for creating data objects is modified so as to selectively inhibit both the allocation of memory space for the data object and the initialization of data within the data object. A data object with hidden pointers created by a previous program invocation is retrieved into memory by a subsequent program invocation. The modified object creation process is then advantageously applied to the data object. As a result, the hidden pointers within the data object are initialized without allocating new memory or disturbing the data within the data object. Where the object creation process includes the execution of a user supplied initialization function, the initialization function is modified so as to inhibit the initialization of data within the data object. A global flag is set by the object creation process and tested within the initialization function.

Abstract: A method and apparatus for providing a differential microphone with a desired frequency response are disclosed. The desired frequency response is provided by operation of a filter, having an adjustable frequency response, coupled to the microphone. The frequency response of the filter is set by operation of a controller, also coupled to the microphone, based on signals received from the microphone. The desired frequency response may be determined based upon the distance between the microphone and a source of sound, and may comprise both a relative frequency response and absolute output level. The frequency response of the filter may comprise the substantial inverse of the frequency response of the microphone to provide a flat response. Furthermore, the filter may comprise a Butterworth filter.

Abstract: A system pattern-based speech recognition, e.g., a hidden Markov model (HMM) based speech recognizer using Viterbi scoring. The principle of minimum recognition error rate is applied by the present invention using discriminative training. Various issues related to the special structure of HMMs are presented. Parameter update expressions for HMMs are provided.

Abstract: A speech coding system robust to frame erasure (or packet loss) is described. Illustrative embodiments are directed to a modified version of CCITT standard G.728. In the event of frame erasure, vectors of an excitation signal are synthesized based on previously stored excitation signal vectors generated during non-erased frames. This synthesis differs for voiced and non-voiced speech. During erased frames, linear prediction filter coefficients are synthesized as a weighted extrapolation of a set of linear prediction filter coefficients determined during non-erased frames. The weighting factor is a number less than 1. This weighting accomplishes a bandwidth-expansion of peaks in the frequency response of a linear predictive filter. Computational complexity during erased frames is reduced through the elimination of certain computations needed during non-erased frames only.

Abstract: A method of identifying untestable faults in a logic circuit. A lead in the circuit is selected and the circuit is analyzed to determine which faults would be untestable if the selected circuit lead were unable to assume a logic 0 and which faults would be untestable if the selected circuit lead were unable to assume a logic 1. Faults that would be untestable in both (hypothetical) cases are identified as untestable faults. Faults which would be untestable if the selected lead were unable to assume a given value may be determined based on an implication procedure. The implication procedure comprises the forward propagation of uncontrollability indicators and the backward propagation of unobservability indicators. An uncontrollability indicator for the given value is assigned to the selected circuit lead and propagated forward through the circuit according to a set of well-defined propagation rules.

Abstract: A method of identifying redundant and untestable faults in a sequential logic circuit. A lead in the circuit is selected and the circuit is analyzed to determine which faults would be hypothetically undetectable at a given time frame if the selected circuit lead were unable to assume a logic 0 at a starting time frame, and which faults would be hypothetically undetectable at the given time frame if the selected circuit lead were unable to assume a logic 1 at the starting time frame. Faults that would be undetectable at the given time frame in both hypothetical cases are identified as redundant and untestable faults. This analysis may be repeated for each of a plurality of time frames in a range of time frames which includes the starting time frame.

Abstract: An automated method for modifying a speech signal communicated over a telephone network by applying a gain factor in a telephone set before any side tone is added thereto. The gain factor is a function of a received signal indicative of the background noise at the location of the telephone set. The gain applied may be a function of both the background noise level and the original speech signal. Either a linear or a non-linear (e.g., compressed) amplification of the original speech signal may be performed, where a compressed amplification results in the higher level portions of the speech signal being amplified by a smaller gain factor than lower level portions. The speech signal may be separated into a plurality of subbands, each resultant subband signal being individually modified in accordance with the present invention.