Abstract: A fast affine projection adaptive filter capable of frequent parameter updating and fast convergence with low complexity. The invention mitigates affine projection adaptive filter complexity due to matrix inversion and the fact that a given excitation vector is weighted and summed into the adaptive coefficient vector many times. Embodiments of the invention achieve fast convergence through sample by sample updating and the exploitation of the shift invariant property of the excitation signal. Complexity is further reduced by a vector update procedure in which a given excitation vector is weighted and added into the coefficient estimate only once. Three illustrative embodiments are presented. The first of these is a relaxed adaptive filter, the second is an adaptive filter without relaxation and a third is an adaptive filter with suspended adaptation. The embodiments are presented in the context of electrical and acoustic echo cancellation.

Abstract: A technique for coding color components of video signals wherein a color component of an image signal sequence is filtered into a plurality of subband image signals. One or more pixels of a first subband image signal are coded using one or more bits of a first quantity of bits. The first subband image signal reflects an image structure of a color component of the image signal sequence. The first subband includes low spatial and temporal frequencies of the color component. A second quantity of bits is used for coding one or more pixels of a second subband image signal. This second quantity of bits comprises one or more bits of the first quantity of bits not used in coding pixels of the first subband image signal. One or more pixels of the second subband image signal are coded using one or more of the second quantity of bits. The second subband image signal reflects image object motion information of the color component of the image signal sequence.

Abstract: An electret transducer array and fabrication technique are disclosed. The array comprises a foil having a layer of insulating material and a layer of metal in contact therewith. The layer of metal comprises one or more discrete areas of metal which define the active areas of one or more transducers in the array. Electrical leads are coupled to the discrete areas of metal. By means of these leads, electrical signals produced by each transducer in response to incident acoustic signals may be accessed. The areas of metal may be formed by selectively removing metal from the foil, or by selective metal deposition. The layer of insulating material is electrostatically charged. The electret transducer array further comprises a porous backplate of sintered metal. The backplate further comprises a rough surface in contact with the layer of insulating material. The backplate serves as a common electrode for transducers of the array.

Abstract: A signal processing arrangement uses a codebook of first vector quantized speech feature signals formed responsive to a large collection of speech feature signals. The codebook is altered by combining the first speech feature signals of the codebook with second speech feature signals generated responsive to later input speech patterns during normal speech processing. A speaker recognition template can be updated in this fashion to take account of change which may occur in the voice and speaking characteristics of a known speaker.

Abstract: A geometric vector quantizer coding technique is illustrated in the context of a full motion video coder based on a three-dimensional sub-band framework. The original image data is decomposed into different spatial-temporal frequency bands and based on the data in each band, different quantization strategies are applied to the bands. The lowest spatial-temporal frequency band is coded using a traditional three-dimensional switched predictor and optimum scaler quantizer. The non-dominant sub-bands are coded using the vector quantization approach to efficiently encode the images while appropriately exploiting the sparse, highly structured nature of the data to design the codebooks. Variable resolution is achieved using these techniques and no training is needed to establish or maintain the codebook. A fast codebook search technique is provided for identifying the optimal codebook vector for representing a block of input data. Examples of two and three level geometric vector quantizers are also provided.

Abstract: Branchwood signals are generated by a tailbiting convolutional coder. The receiver stores received branchword signals in memory such that the branchword signals are accessible in a logically circular sequence. The receiver then performs Viterbi updates on the logically circular sequence of branchwords, the sequence comprising more than N branchwords. The receiver stops performing Viterbi updates in response to an indication that update decisions have become repetitive. A set of generated decision vectors resulting from the Viterbi updates is then modified in response to the indication. A decoded signal is generated by performing a Viterbi traceback procedure using the modified set of decision vectors. The indication that Viterbi update decisions have become repetitive may be provided by a predetermined fixed number of Viterbi updates, the number reflecting an estimate of when said update decisions will likely become repetitive.

Abstract: The present invention provides a method of coding a set of multi-level signals, such as video signals. The method comprises filtering the set of multi-level signals into one or more sets of multi-level sub-band signals; allocating a first quantity of bits for use in coding a first set of sub-band signals; coding the first set of multi-level sub-band signals using zero or more of the first quantity of allocated bits; allocating a second quantity of bits for use in coding a second set of sub-band signals, the second quantity of bits comprising zero or more of the first quantity of bits not used in coding the first set of multi-level sub-band signals; and coding the second set of multi-level sub-band signals using zero or more of the second quantity of allocated bits. Three examples of the method are provided. A hardware implementation involving the use of digital signal processors operating in parallel is also provided.

Abstract: An image-processing system for perceptual coding of images by analyzing one or more images to determine levels of imperceptible noise and encoding a first image to produce encoded values without introducing noise exceeding a determined level of imperceptible noise is disclosed. Analyzing of images is carried out under a set of user determined conditions comprising viewing distance and lighting. The encoded image is communicated and decoded to produce representation of the first image for display. As part of displaying the representation of the first image, a halftoning process is carried out. The halftoning process may comprise the interpolation of data values of the representation of the first image as well as the addition of micro-dither to the representation. Illustrative halftoning techniques include classical halftoning, least-squares halftoning, and error diffusion halftoning. Printer model-based halftoning techniques for both least-squares halftoning and error diffusion halftoning are presented.

Abstract: The invention provides a method and apparatus for transmitting digital signal information to a receiver using a plurality of antennas. The invention involves applying a channel code to a digital signal producing one or more symbols. A plurality of symbol copies is made and each copy is weighted by a distinct time varying function. Each antenna transmits a signal based on one of the weighted symbol copies. Any channel code may be used with the invention, such as a convolutional channel code or block channel code. Weighting provided to symbol copies may involve application of an amplitude gain, phase shift, or both. The present invention may be used in combination with either or both conventional interleavers and constellation mappers.

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: An invention for providing diversity for direct sequence spread spectrum wireless communication systems is presented. The invention provides a transmitting technique for communicating a first signal comprising one or more segments to a receiver with use of a plurality of M antennas. The first signal includes one or more signal segments. The technique comprises forming M copies of the first signal; for a segment of a signal copy, weighting each of two or more sub-segments of the segment with a distinct signal, wherein a sequence of the distinct weighting signals for the segment is distinct from sequences of signals weighting the same segment of one or more other signal copies; and for each of M weighted signal copies, transmitting a signal to the receiver using a distinct antenna, the transmitted signal based on the weighted signal copy. Illustratively, the first signal is a spread spectrum signal. Moreover, the process of weighting comprises applying a phase shift to a sub-segment.

Abstract: A method for converting phonetic Chinese into Chinese characters effectuated by ascertaining an optimal path through a lattice of possible combinations of Chinese characters by calculating the probability of adjacent Chinese characters appearing in Chinese character text.

Abstract: A method and apparatus for processing a reconstructed speech signal from an analysis-by-synthesis decoder are provided to improve the quality of reconstructed speech. By operation of the invention, one or more traces in a reconstructed speech signal are identified. Traces are sequences of like-features in the reconstructed speech signal. The like-features are identified by time-distance data received from the long term predictor of the decoder. The identified traces are smoothed by one of the known smoothing techniques. A smoothed version of the reconstructed speech signal is formed by combining one or more of the smoothed traces. The original reconstructed speech signal may be that provided by a long term predictor of the decoder. Values of the reconstructed speech signal and smoothed speech signal may be combined based on a measure of periodicity in speech.

Abstract: A method and apparatus for improving the performance of a phase-locked loop are disclosed. A phase error between two signals is sensed and a temporary increase in the bandwidth of the phase-locked loop is provided responsive to the sensed phase error. The phase-locked loop may comprise a charge pump and the temporary increase in the bandwidth of the phase-locked loop may comprise a temporary increase in charge pump current. An increase in phase-locked loop bandwidth is followed by a decrease in the bandwidth responsive to a decrease in phase error. The decrease in bandwidth may comprise a decrease in charge pump current.