Abstract: A method and apparatus are described for application to an adaptive equalizer, or filter, whereby complex valued coefficients for such an equalizer may be updated with a significant reduction in complexity as compared to prior art methods. Specifically, the method of the invention causes a sum and a difference to be determined for a real-valued and an imaginary-valued component of an input signal sample to the equalizer. Those sum and difference terms are then applied according to the method of the invention to a determination of an update term for a coefficient under consideration. Because the method of the invention results in a material reduction in the number of addition/subtraction operations required for the coefficient update process, the hardware infrastructure needed to implement that process is considerably simplified.

Abstract: A device for providing conferencing communications. The device comprises: a summing circuit for forming a signal sum equal to a sum of signals received from at least three information signal sources along with an echo compensation signal, the signal received from each information source including echoes and the signal sum including an aggregation of such echoes; an adaptive filter having an input and an output, the input being solely responsive to the signal sum and the output being only coupled to the summing circuit, the filter forming the echo compensation signal which is an estimate of the aggregation of echoes; and a low level training sequence generator having an amplitude less than a predetermined level.

Abstract: Disclosed are systems, methods and articles of manufacture for performing high resolution N-best string hypothesization during speech recognition. A received input signal, representing a speech utterance, is processed utilizing a plurality of recognition models to generate one or more string hypotheses of the received input signal. The plurality of recognition models preferably include one or more inter-word context dependent models and one or more language models. A forward partial path map is produced according to the allophonic specifications of at least one of the inter-word context dependent models and the language models. The forward partial path map is traversed in the backward direction as a function of the allophonic specifications to generate the one or more string hypotheses. One or more of the recognition models may represent one phone words.

Abstract: The breakdown of an ultra-thin dielectric layer is detected by applying a test signal to the layer. Measurements are taken of noise signals present in the layer during the application of the test signal. At breakdown, a significant increase occurs in the amplitude of the measured noise signals.

Abstract: Applicants have discovered that the intermodal beat noise of a fiber multimode laser can be substantially reduced by providing the fiber with an output coupler of broadened bandwidth. In a preferred embodiment, a reduced-noise, high power light source comprises a cladding pumped fiber laser having a chirped output grating. Experiments show that increasing the output bandwidth from 0.254 to 0.577 nm reduces the relative intensity-to-noise ratio 10 dB in a Nd-doped fiber laser. Increasing the bandwidth from 0.2 nm to 0.3 nm in a Yb-doped laser similarly reduces the noise by 12 dB.

Abstract: The method of making a metallized magnetic substrate for devices including a magnetic component involves providing an unfired ceramic body. In one exemplary embodiment, the method further involves making one or more vias through the ceramic body, coating the via side walls with conductive material, forming an aperture through the ceramic body, such that an aperture edge intersects the via, and metallizing the unfired ceramic body such that a conductive pathway is formed that includes the conductive material in the via. Finally, the metallized unfired ceramic body is fired in conventional fashion, optionally followed by deposition of additional conductor material.

Abstract: A microstrip patch filter in which a dielectric has a ground plane printed on one of its faces and a conductive arrangement printed on the other of said faces, the conductive arrangement includes a flat patch, input and output leads electromagnetically coupled to the flat patch, the flat patch or the dielectric substrate has a reactance-enhancing metallic constriction located along a portion of the patch. When the constriction is in the patch it forms a current-concentrating inductive constriction. When the constriction is in the dielectric substrate, it enhances the capacitance. In an embodiment, the patch has two mutually-transverse constrictions that divide the patch into four sub-patches cross-connected by current-concentrating inductive constriction.

Abstract: Determining by a computer mouse the pressure that an operator uses to grip the mouse and to actuate switches on the mouse and alerting the operator by the mouse if the pressure is excessive. This monitoring and alerting gives the operator immediate and consistent feedback on misuse of the mouse. In a first embodiment, the computer mouse alerts the operator to excessive pressure by means of an audio alarm. In the second embodiment, the computer mouse transmits the pressure information to a connected computer. The computer can alert the operator with an audio message or alert the operator by displaying the alerting information on a display screen of the computer.

Abstract: An embodiment of the present invention locates facial features in an image by bandpass filtering the image and then performing morphological operations followed by a thresholding operation. This initial processing identifies candidate areas where facial features may be located. The candidate areas are evaluated by classifiers to determine if a facial feature, such as an eye or mouth, has been located.

Abstract: A method for automatically converting a decision tree into one or more weighted finite-state transducers. Specifically, the method in accordance with an illustrative embodiment of the present invention processes one or more terminal (i.e., leaf) nodes of a given decision tree to generate one or more corresponding weighted rewrite rules. Then, these weighted rewrite rules are processed to generate weighted finite-state transducers corresponding to the one or more terminal nodes of the decision tree. In this manner, decision trees may be advantageously compiled into weighted finite-state transducers, and these transducers may then be used directly in various speech and natural language processing systems. The weighted rewrite rules employed herein comprise an extension of conventional rewrite rules, familiar to those skilled in the art.

Abstract: A uniform building entrance protector housing construction includes a base layer in the shape of a generally rectangular box with a back wall and a side wall extending from the perimeter of the back wall. A number of equipment mounts inside the base layer are arranged to support wire cable equipment when mounted in the base layer, and first base layer hinge parts are provided on a section of the side wall. The housing construction also has a top layer including a frame portion with a frame wall having a top opening and a bottom opening. A first frame wall section has first frame portion hinge parts located on the section to engage the first base layer hinge parts for swinging movement about a first hinge axis. A second frame wall section has second frame portion hinge parts. The top layer also has a cover member with cover hinge parts located to engage the second frame portion hinge parts for swinging movement about a second hinge axis.

Abstract: A system for, and method of operation of, estimating a blocking probability for at least a portion of a network and a model of the same are provided. The blocking probability represents a likelihood that a transmitted signal will arrive at least at the portion of the network. The system includes an estimator controller and a processing controller. The estimator controller derives both a direct and an indirect estimator. The direct controller is derived as a function of a number of losses and arrivals occurring with respect to at least the portion of the network during a period of time, and the indirect estimator is derived as a function of an offered load with respect to at least the portion of the network during the period of time. The processing controller, that is associated with the estimator controller, applies a weighting factor to the direct and indirect estimators to derive the blocking probability. The weighting factor minimizes variance of combined estimates.

Abstract: The modulation scheme has two versions that can transform two data streams into two mutually orthogonal modulated signals within the same carrier-frequency band. In one embodiment, the modulation scheme is a quadrature-phase shift-keying (QPSK) modulation scheme having an in-phase part and a quadri-phase part. In this embodiment, one version of the QPSK modulation scheme conforms to the IS-95 telecommunications standard. In this version, the in-phase part is applied to a first data stream to generate an in-phase signal that is multiplied by a cosine function of the carrier frequency. In a parallel path, the quadri-phase part is applied to the first data stream to generate a quadri-phase signal that is multiplied by a sine function of the carrier frequency. The resulting signals are combined to generate a conventional IS-95 signal.

Abstract: Directory assistance (106) is used to train a voice-dialing system (107) as a part of normal voice-dialing activities, thereby avoiding the need for a separate training session. The voice-dialing system connects a caller to directory assistance when a voice-dialing attempt by the caller does not match any voice tags in the caller's voice-dialing directory (130) and the caller signals a desire to use directory assistance. The voice-dialing system monitors the call to directory assistance, and captures and stores the telephone number provided by directory assistance to the caller. After the directory assistance call is completed, the voice-dialing system asks the caller to repeat the voice tag for the telephone number, forms a data representation of the voice tag from multiple received instances of the voice tag, records the data representation of the voice tag along with the captured telephone number in the caller's voice-dialing directory, and at the caller's option initiates a call to the telephone number.

Abstract: A scheme for variable-delay instructions in a digital processor that allows for variable delay of some instructions to increase performance at different frequencies. The variable-delay (VD) feature allows flag-modifying instructions to execute in a differing number (1 or 2) of clock cycles, depending on the application. In applications that clock the processor at less than maximum frequency, instructions that modify the flag are executed in one clock cycle. In applications that clock the processor at its maximum frequency, the instructions that modify the flag are executed in two clock cycles. If the critical path, and consequently the maximum frequency, of a processor is determined by a flag-modifying operation immediately followed by a flag-reading operation, then the VD scheme helps increase performance at either frequency. The performance increase is proportional to the difference in delays between the critical path associated with flag-modifying and other critical paths.

Abstract: An apparatus is provided for adding a dither signal to an optical carrier modulated with an information signal. The apparatus includes a dithering element having an input and an output in series with an optical carrier modulated with an information signal. A differential element having an inverting input, a non-inverting input, and an output is coupled to the input of the dithering element. A signal generator, which is provided for producing a desired dither signal to be added to the modulated optical carrier, is coupled to the non-inverting input of the differential element. A feedback element produces a signal related to the output of the dithering element and couples the signal related to the output of the dithering element to the inverting input of the differential element.

Abstract: A system for generating aural announcements of incoming calls within a premises has a premises phone connected within a premises phone line that forms first and second communication channels. The first communication channel is adapted for connection and communication to a phone network. An announcement controller is connected within the phone line and generates aural announcement signals along the second communication channel to a speaker which plays corresponding aural announcements. A switch may be operatively connected between the premises phone and announcement controller and switches the announcement controller into the second communication channel for transmitting aural announcement signals to a premises phone via the switch.

Abstract: An integrated circuit including a circuit for improved efficiency of internal data transfer comprises: a processor core having a buffer memory; a random access memory having a read and write cycle time of a one clock cycle, the random access memory comprising a memory array with a predetermined word width and a data latch coupled to the memory array; a bi-directional data bus coupling the processor core to the random access memory, the bi-directional data bus having a data width which is a multiple of at least one times the predetermined word width; and, a signal circuit coupled to the data latch wherein the data latch is responsive to the signal circuit to latch data from the bi-directional data bus prior to writing the data to the memory array, wherein alternately reading two consecutive data words and writing two consecutive words occurs on an average in the clock cycle.

Abstract: A method and apparatus for recognition of objects such as faces in images using signal compression techniques (i.e., coding techniques) in which a portion of the image which includes the object to be recognized (e.g., the face) is coded, and the resultant coded data is matched against previously coded and stored training data which makes up a known object database. A given object in an input image signal is matched to one of a plurality of known objects stored in a database, wherein the stored representation of each of the known objects comprises a codebook generated based on training image signals comprising the known object. A first illustrative embodiment comprises the steps of decomposing the given object into blocks; performing a plurality of encodings of the given object, each encoding comprising coding the object with use of one of the codebooks; determining a coding error for each encoding; and matching the given object to one of the known objects based on the coding errors.

Abstract: A method and apparatus for obtaining a soft symbol decoded output of a received signal by a two pass Viterbi operation. The technique is especially advantageous where the signal is convolutionally encoded with large constraint lengths. During the first pass, the error-correction co-processor (ECCP) is programmed for hard decoded output alone. After all the received symbol sets are hard-bit decoded, a second pass Viterbi operation is performed. Using the previously decoded hard bit to identify the most likely next state at an initial time instant, and initializing the present states at that initial time instant with pre-saved accumulated costs from the first pass Viterbi operation, branch metrics are computed for those state transitions leading to the most likely next state at that time instant.