3153 results about "Weight coefficient" patented technology

The present invention discloses a distributed base station serving mobile stations, which comprises multiple antenna-RF processing units and a central processing unit remotely connected with said multiple antenna-RF processing units. Each of the antenna-RF processing units has an antenna array formed by multiple antennas. Said central processing unit comprises: an MIMO coding unit for MIMO coding of data of each of the mobile stations so that data of each of the mobile stations generates multi-path data symbols; a beam-forming unit for multiplying each path data symbol of the multi-path data symbols of each of the mobile stations by a weighted coefficient vector corresponding to this path data symbol, to obtain beam-forming data of each path data symbol of each of the mobile stations; and an overlapping unit for, with respect to each antenna in said antenna-RF processing units, overlapping beam-forming data of the respective mobile stations for this antenna and feeding the overlapped beam-forming data to this antenna for transmission. The present invention further discloses a communication system having the above distributed base station and the corresponding signal transmission method.

A graphics system including a custom graphics and audio processor produces exciting 2D and 3D graphics and surround sound. The system includes a graphics and audio processor including a 3D graphics pipeline and an audio digital signal processor. The system achieves highly efficient full-scene anti-aliasing by implementing a programmable-location super-sampling arrangement and using a selectable-weight vertical-pixel support area blending filter. For a 2×2 pixel group (quad), the locations of three samples within each super-sampled pixel are individually selectable. A twelve-bit multi-sample coverage mask is used to determine which of twelve samples within a pixel quad are enabled based on the portions of each pixel occupied by a primitive fragment and any pre-computed z-buffering. Each super-sampled pixel is filtered during a copy-out operation from a local memory to an external frame buffer using a pixel blending filter arrangement that combines seven samples from three vertically arranged pixels. Three samples are taken from the current pixel, two samples are taken from a pixel immediately above the current pixel and two samples are taken from a pixel immediately below the current pixel. A weighted average is then computed based on the enabled samples to determine the final color for the pixel. The weight coefficients used in the blending filter are also individually programmable. De-flickering of thin one-pixel tall horizontal lines for interlaced video displays is also accomplished by using the pixel blending filter to blend color samples from pixels in alternate scan lines.

Methods and apparatus, including computer program products, for identifying matches between disparate schemas calculates a degree of similarity between elements of two schemas using each of multiple matching processes. The calculated degrees of similarity are combined using a first weighting vector to produce first combined degrees of similarity. The first weighting vector includes multiple weighting coefficients and each weighting coefficient corresponds to one of the matching processes. The weighting coefficients are tuned using information relating to a predicted degree of matching accuracy associated with the first weighting vector.

A MIMO-OFDM communication method for transmitting and receiving data using MIMO-OFDM communication has, on the transmitting side, a step of performing precoding processing of transmission data using a prescribed precoding method, and a step of performing OFDM modulation of the precoding-processed data and transmitting the data from a plurality of transmission antennas, and has, on the receiving side, a step of performing MIMO decoding processing of reception signals received by a plurality of reception antennas according to a ZF decoding algorithm or an MMSE decoding algorithm, and outputting the data as a transmission data stream, a step of calculating weighting coefficients according to the precoding method, and a step of performing weighted decoding processing of the transmission data obtained by the MIMO decoding processing, by multiplying path metric of the decoding processing by the weighting coefficients.

The invention provides a remote sensing image registration method of a multi-source sensor, relating to an image processing technology. The remote sensing image registration method comprises the following steps of: respectively carrying out scale-invariant feature transform (SIFT) on a reference image and a registration image, extracting feature points, calculating the nearest Euclidean distances and the nearer Euclidean distances of the feature points in the image to be registered and the reference image, and screening an optimal matching point pair according to a ratio; rejecting error registration points through a random consistency sampling algorithm, and screening an original registration point pair; calculating distribution quality parameters of feature point pairs and selecting effective control point parts with uniform distribution according to a feature point weight coefficient; searching an optimal registration point in control points of the image to be registered according to a mutual information assimilation judging criteria, thus obtaining an optimal registration point pair of the control points; and acquiring a geometric deformation parameter of the image to be registered by polynomial parameter transformation, thus realizing the accurate registration of the image to be registered and the reference image. The remote sensing image registration method provided by the invention has the advantages of high calculation speed and high registration precision, and can meet the registration requirements of a multi-sensor, multi-temporal and multi-view remote sensing image.

The invention provides in one aspect methods and apparatus for use with C-arm and other CT systems, e.g., with non-rigid geometries. In such systems, by way of example, calibration can be performed to determine the exact position of the x-ray source and the exact orientation of the detector where each projection measurement is made. Next, a weighting coefficient can be determined for the voxels in each plane of a reconstruction volume at every possible projection. Finally, the order in which to process the voxels during image reconstruction can be determined. Following an actual CT scan procedure in which scans are obtained of a volume to be constructed, a system according to these and related aspects of the invention can use an optimal, pre-calculated processing method, while utilizing offsets and weighting coefficients determined during calibration, for performing backprojection image reconstruction.

A method for carrying out channel equalization in a radio receiver wherein an impulse response is estimated, noise power is determined by estimating a co-variance matrix of the noise contained in a received signal before prefiltering, and tap coefficients of prefilters and an equalizer are calculated. The method comprises determining the noise power after prefiltering by estimating a noise covariance matrix, after which input signals of the channel equalizer are weighted by weighting coefficients obtained from the noise covariance estimation.

A new scheduling discipline is disclosed for the forward link of a wireless network such as a CDMA network. The new discipline, which we denominate Modified LWDF (M-LWDF), resembles the known LWDF discipline in that only one queue is served at a time, and the queue to be served is that having the largest weighted delay. According to M-LWDF, the weighted delay of the i'th queue is defined aswherein Wi(t) is a packet delay, a queue length, or an increasing function of packet delay or queue length for the i'th queue, ci(t) is a weight coefficient descriptive of channel conditions for the i'th user, and gammai is a fixed constant that can be chosen arbitrarily. We have found that if any scheduling discipline can yield stable queues for a given traffic pattern of arriving packets, then M-LWDF can, even under changing channel conditions. Moreover, we have found that M-LWDF can provide favorable QoS in terms of delay bounds.

The invention discloses a driving behavior analysis method and a driving behavior analysis system. The driving behavior analysis method comprises the following steps: collecting all bad driving behavior data of a driver when the driver drives; uploading the bad driving behaviors data to a data analysis center through networks; multiplying times of various bad driving behaviors and corresponding weight coefficients of bad driving behaviors in the data analysis center to obtain a corresponding weight value; obtaining an assessed value by subtracting the sum of all weight values of bad driving behaviors during a driving from a preset full score of good behaviors. The driving behavior is good if the percentage of the assessed value and the full score value is greater than or equal to X, is general if the percent is greater than or equal to Y and less than X, and is bad if the percent is greater than or equal to Z and less than Y and fails if the percent is less than Z. The analysis method and the analysis system of the driving behaviors can be used for comprehensively, scientifically and accurately assessing bad driving behaviors of a driver and warning the driver in time and meanwhile a company can manage drivers conveniently.

The invention relates to a vehicle flow predicting method based on an integrated LSTM neural network. On the basis of historical data obtained by vehicle flow detection, an integrated LSTM neural network vehicle flow prediction model is established to carry out vehicle flow prediction, so that the generalization error of the prediction model is reduced and the accuracy is improved. The method comprises the following steps that: data preprocessing is carried out; according to a preprocessed vehicle flow time sequence value, a vehicle flow matrix data set is constructed and the vehicle flow of an (n+1)th period of time is predicted by using first n periods of time, wherein each period of time is delta t expressing the time length and the unit is min; a plurality of different LSTM neural network models are constructed by using different initial weights; on the basis of a bagging integrated learning method, a training set and a verification set are constructed; a plurality of LSTM neural networks are trained to obtain an optimized module; a weighting coefficient of the single LSTM model is calculated by using the verification set; and inverse transformation and reverse normalization are carried out on a predicted vehicle flow value to obtain a predicted vehicle flow and integrated weighting is carried out to obtain a vehicle flow value predicted finally by the model.

With conventional source separator devices, specific frequency bands are significantly reduced in environments where dispersed static is present that does not come from a particular direction, and as a result, the dispersed static may be filtered irregularly without regard to sound source separation results, giving rise to musical noise. In an embodiment of the present invention, by computing weighting coefficients which are in a complex conjugate relation, for post-spectrum analysis output signals from microphones (10, 11), a beam former unit (3) of a sound source separator device (1) thus carries out a beam former process for attenuating each sound source signal that comes from a region wherein the general direction of a target sound source is included and a region opposite to said region, in a plane that intersects a line segment that joins the two microphones (10, 11). A weighting coefficient computation unit (50) computes a weighting coefficient on the basis of the difference between power spectrum information calculated by power calculation units (40, 41).

The embodiment of the invention provides a model training method and device based on federated learning and a server. The method comprises the steps of sending a to-be-trained model to a plurality ofworking nodes; receiving a local model fed back by the plurality of working nodes, wherein the local model is obtained by training a to-be-trained model by each working node according to own data; determining the precision of each local model according to the test data set; determining a weight coefficient of each local model according to the precision of each local model, wherein the weight coefficient is positively correlated with the precision; and updating the to-be-trained model according to the plurality of local models and the corresponding weight coefficients. According to the method provided by the embodiment of the invention, by increasing the weight coefficient of the high-precision local model and reducing the weight coefficient of the low-precision local model, the convergencerate of the to-be-trained model is increased, and the precision of the to-be-trained model is improved.

A plurality of ECG Superscore formulae, created from multiple parameter ECG measurements including those from advanced ECG techniques, can be optimized using additive multivariate statistical models or pattern recognition procedures, with the results compared against a large database of ECG measurements from individuals with known cardiac conditions and/or previous cardiac events. Superscore formulae utilize multiple ECG parameters and accompanying weighting coefficients and allow data obtained from any given patient to be used in calculating that patient's ECG Superscore results. ECG Superscores have retrospectively optimized accuracy for identifying and screening individuals for underlying heart disease and/or for determining the risk of future cardiac events. They thus have greater predictive value than that of any conventional or advanced ECG measurement alone or of any non-optimized combinations of conventional or advanced ECG measurements that have been used in the past. Ongoing optimization of ECG Superscore diagnostic and predictive accuracy may be realized through the iterative adjustment of Superscore formulae based on the incorporation of data from new patients into the database and/or from longitudinal follow-up of the disease and cardiac event status of existing patients.

An optical transmission network includes an optical transmitter photonic integrated circuit (TxPIC) chip, utilized in an optical transmitter and has a plurality of monolithic modulated sources integrated for multiple signal channels on the same semiconductor chip is provided with channel equalization at the optical receiver side of the network that permits one or more such integrated modulated sources in the TxPIC chip to be out of specification thereby increasing the chip yield and reducing manufacturing costs in the deployment of such TxPIC chips. FEC error counts at the FEC decoder on the optical receiver side of the network includes counters that accumulate a plurality of bit pattern-dependent error counts based on different N-bit patterns in the received data bit stream. The accumulated counts of different N-bit patterns are utilized to provide for corrections to threshold and phase relative to the bit eye pattern as well as provided for weight coefficients for the optical receiver equalization system. The deployment of this type of equalization in a digital OEO REGEN network substantially reduces, if not eliminates, the need for dispersion compensating fiber (DCF) or EDFAs in an optical link of the network and enhances the optical receiver tolerance to chromatic dispersion (CD) so that an increase in chip yield is realized for TxPIC chips not operating with acceptable operational parameters, particularly with a desired frequency chirp parameter relative to at least one of the TxPIC modulated sources.

A picture signal encoding apparatus on which, even if a picture is not uniform in picture pattern, the picture can be rendered substantially uniform in picture quality, the volume of bits occupying a virtual buffer is stabilized, deterioration in picture quality is rendered less visible and in which stable rate control is assured. The picture signal encoding apparatus, dividing a picture making up moving picture signals into plural macro-blocks MB and encoding the signals from one macro-block to another, includes a relative coding difficulty calculating circuit for calculating the relative coding difficulty from one macro-block to another MB, and a characteristic value calculating circuit for calculating the weighting coefficients (characteristic values) specifying visibility degree of picture quality deterioration from one macro-block MB to another. The picture signal encoding apparatus also includes a bit rate controller for calculating the quantization scale for encoding each macro-block MB using the complexity degree and the volume of allocated bits for macro-block-based encoding.

In performing SVD-MIMO transmission, a set-up procedure is simplified while assuring a satisfactory decoding capability with a reduced number of antennas. A transmitter estimates channel information based on reference signals sent from a receiver, determines a transmit antenna weighting coefficient matrix based on the channel information, calculates a weight to be assigned to each of components of a multiplexed signal, and sends, to the receiver, training signals for respective signal components, the training signals being weighted by the calculated weights. On the other hand, the receiver determines a receive antenna weighting coefficient matrix based on the received training signals.

The invention relates to a time-space domain hybrid video noise reduction device and a time-space domain hybrid video noise reduction method, and belongs to the technical field of video image processing. The device comprises a time domain noise reduction module, a smoothing coefficient storage module, a reference frame storage module and a space domain noise reduction module. The method comprises the following steps that: the time domain noise reduction module calculates noise variance, the motion intensity of a current point and a weighting coefficient of the current point according to the current point and a reference point; the time domain noise reduction module calculates and stores a smoothing coefficient of the current point and a time domain filtering result of a current frame respectively; and the space domain noise reduction module performs space domain filtering to obtain noise-reduced image data. Therefore, the combined application of time domain noise reduction and space domain noise reduction is realized, a good noise reduction effect can be achieved, system overhead required by the noise reduction can be effectively reduced, and system cost is lowered. The noise reduction device is simple in structure; and the noise reduction method can be conveniently and widely applied.

The invention provides an adaptive feature fusion target tracking method. The adaptive feature fusion target tracking method comprises in a first frame image, initializing a target area, and constructing a location filter and a scaling filter; extracting detection samples around a target, respectively calculating an HOG (Histogram of Oriented Gradient) feature and a CN (Color Name) feature, and obtaining a response value through the location filter; calculating the feature weight according to the response value, normalizing the weight coefficient, fusing the feature response value, and selecting the point having the largest response value as the central location of the target; judging whether a shielding phenomenon occurs according to the target response, under the shielding condition, only updating the scaling filter without updating the target location filter, carrying out a circular process, and obtaining the target location of each frame. The adaptive feature fusion target trackingmethod is advantaged in that an adaptive feature fusion method is provided, a model updating strategy based on APCE (Average Peak-to-Correlation Energy) is designed, and the target tracking precisionand the robustness under the shielding condition can be greatly improved.