149results about How to "Increase computational cost" patented technology

A method for padding, filtering, denoising, image enhancing and increased time-frequency acquisition is described for digitized data of a data set where unknown data is estimated using real data by adding unknown data points in a manner that the padding routine can estimate the interior data set including known and unknown data to a given accuracy on the known data points. The method also provides filtering using non-interpolating, well-tempered distributed approximating functional (NIDAF)-low-band-pass filters. The method also provides for symmetric and/or anti-symmetric extension of the data set so that the data set may be better refined and can be filtered by Fourier and other type of low frequency or harmonic filters.

The present invention provides a display apparatus which enables high-accuracy, hands-free operation, while adopting a low-cost configuration. The display apparatus according to the present invention includes: an obtainment unit that obtains a display element which is information to be displayed to the user; a determining unit that determines a display position of the obtained display element, at which a portion of the obtained display element can not be seen by the user; a display unit that displays the display element at the determined display position; a detection unit that detects a direction of change when an orientation of the user's head is changed; a selection unit that selects the display element to be displayed in the detected direction of change; and an execution unit that executes a predetermined process related to the selected display element.

A computer system having an optimized system for data integrity verification is disclosed. The computer system includes a processor and memory storage that is accessible by the processor. The memory storage includes stored data that is subject to data verification. The system also includes data verification software that is executable by the processor to verify the integrity of the stored data. The data verification software is executable to conduct a first integrity check on a given set of stored data and, if an error is detected, to conduct a second integrity check on the given set of stored data. The second integrity check may have a larger computational cost than the first integrity check thereby increasing the overall efficiency and response time of the system to data verification events.

The present invention relates to a method for determining drive values for driving a lighting device at a desired brightness and color. The method comprising the steps of determining a first luminous flux weight ratio based on the desired color and a first drive current for driving each of the differently colored LEDs, determining a first luminous flux for each of the differently colored LEDs based on the desired brightness and the first luminous flux weight ratio, comparing, for each of the differently colored LEDs, the first luminous flux with a nominal luminous flux for a plurality of different drive currents, selecting, for each of the differently colored LEDs, a preferred drive current that at least can produce the first luminous flux, determining a second luminous flux weight ratio based on the desired color and the selected drive currents for each of the differently colored LEDs, determining a second luminous flux for each of the differently colored LEDs based on the desired brightness and the second luminous flux weight ratio, and determining a duty cycle for each of the differently colored LEDs at the selected drive currents, wherein the selected currents at the determined duty cycles produces the second luminous flux for each of the differently colored LEDs. The present invention provides for the possibility to limit the number of necessary computational steps for determining preferred drive currents. Furthermore, an increase in number of current level and/or differently colored LEDs would only slightly increase the computational cost.

The application describes a method and an apparatus to prevent clipping of an audio signal when protection against signal clipping by received audio metadata is not guaranteed. The method may be used to prevent clipping for the case of downmixing a multichannel signal to a stereo audio signal. According to the method, it is determined whether first gain values (4) based on received audio metadata are sufficient for protection against clipping of the audio signal. The audio metadata is embedded in a first audio stream (1). In case a first gain value (4) is not sufficient for protection, the respective first gain value (4) is replaced with a gain value sufficient for protection against clipping of the audio signal. Preferably, in case no metadata related to dynamic range control is present in the first audio stream (1), the method may add gain values sufficient for protection against signal clipping.

This invention relates to a system for estimating the position, speed and orientation of a vehicle (10), comprising means for determining the components of two noncollinear constant unit vectors _b, _b according to vehicle body axes; means for determining the components of said noncollinear constant unit vectors {right arrow over (g)}_t, {right arrow over (e)}_t according to Earth's axes; means for determining the three components of angular velocity _b of the vehicle in body axes; means for correcting said angular velocity _b with a correction u_ω and obtaining a corrected angular velocity {circumflex over (ω)}_b=_b+u_ω; a control module (14) implementing a control law to calculate said correction u_ω, where said control law is:

u_ω=σ(_b×ĝ_b+<?img id="custom-character-00007" he="3.13mm" wi="1.78mm" file="US20070213889A1-20070913-P00901.TIF" alt="custom character" img-content="character" img-format="tif" ?>_b×ê_b)  [1]

where σ is a positive scalar,

such that upon using said corrected angular velocity {circumflex over (ω)}_b=_b+u_ω as input to a module for integrating the kinematic equations, the latter are stable in the ISS sense and the error in the estimation of the direction cosine matrix {circumflex over (B)} and of the Euler angles {circumflex over (Φ)} is bounded.

The invention discloses a robot whole-situation path planning method facing an uncertain obstacle environment of mixed terrain and region, aiming to enabling the robot to finish a task by the lowest risk degree under a mixed terrain environment. The robot whole-situation path planning method comprises the following steps of: (1) detecting and constructing a robot working environment model comprising a movement starting point, a target point, static obstacle position and shape, a possible covering region of an uncertain obstacle, and covering terrain type and region; (2) adopting a weighting passable length of a terrain passable degree as a path evaluation performance index I; (3) adopting a path containing degree for describing an uncertain obstacle covering region covering degree as a path evaluation performance index II; (4) constructing a path feasible region according to a penetration degree between static obstacles to realize efficient path feasibility judgment and repairing; (5) extracting an advantage path public segment for automatically generating a feasible path; and (6) aiming at two performance indexes of the steps (3) and (4), adopting a multi-target cultural algorithm to realize path evolution and optimization.

Methods of preventing spam email in an email system include determining whether a sender of email to a receiver of email has paid a required setup price or cost to become a registered sender of email to the receiver. If it is determined that the sender has not paid the required setup cost, then the receiver can reject the sender's email to thereby eliminate the receipt of unwanted spam messages. In order to qualify as a registered sender, the sender undergoes a qualification scheme wherein it pays the required cost, which may be for example monetary or computationally related, and is assigned an encrypted key by the receiver which allows the sender to send to the receiver email authenticated using the unique key associated with the sender and receiver. The receiver has the option of accepting or rejecting email from a sender in this fashion and can change the required setup cost or requirement at any time to ensure that the sender does not try to corrupt the email system once registered.

The invention relates to a query optimization method based on a join index in a data warehouse. The method comprises the following steps: creating a fact table foreign key join index based on a dimension table surrogate key; creating a join index update strategy; when a predicate condition exists on a dimension table, making a predicate judgment after direct access to dimension table record through the join index, and performing calculation through a dynamic bitmap join index by reference to a bitmap join index to generate a join bitmap index; when the selection rate on the dimension table is very low, further compressing a dimension table bitmap in order to raise the cache level of dimension table bitmap calculation; and calculating simple memory address access by only using an array-bitmap data structure with a reference bitmap join index calculation method based on co-processor memory calculation. Through adoption of the query optimization method, the calculation performance of the bitmap index can be effectively increased. Moreover, the query optimization method can be combined with high-performance co-processors such as GPU and Intel Xeon Phi TM to accelerate the index calculation performance, and can be widely applied to the field of data warehouses.

A method of removing ringing and blocking artifacts from a decompressed digital image. In one method, a background value of the digital image a background region and foreground regions is determined, and a threshold value is computed. A mapping of the digital image is then generated by thresholding the digital image based on the threshold value to produce a thresholded image, and enlarging the foreground regions of the thresholded image to form a map image. The mapping includes the background pixels of the map image. The background region of the decompressed digital image is then cleansed based on the mapping. In an alternate method, ringing artifacts are removed from a decompressed digital image based on a signal-to-noise ratio of the image. In yet another alternate embodiment, ringing artifacts are from a decompressed digital color image that includes a chrominance channel and a luminance channel. An alternate embodiment removes ringing and blocking artifacts.

To realize a GDFE precoder for multi-user MIMO systems, which significantly reduces the computational cost while resulting in no capacity loss, one method comprises computing an effective UL channel matrix H_UL using one of two methods H_UL=H_DL^H, or H_UL=[(P_t/N_t)H_DL^HH_DL+I]^−1/2H_DL^H; extracting H_k from H_UL; computing a singular value decomposition of the DL channel between the BS and k^th UT, H_k, for all K UTs, H_k=U_kS_kV_k^H; extracting all singular values as s=[diag(S₁), . . . , diag(S_K)]; extracting a vector ŝ from s by choosing first utmost N_t largest non-zero singular values of s; sorting elements in ŝ in decreasing order; performing water-filling to allocate power and obtain a diagonal matrix Γ_k representing power allocations corresponding to the singular values of the k^th UT; computing an UL covariance matrix for each UT as Φ_k=U_kΓ_kU_k^H; and obtaining an overall input covariance matrix D for the equivalent UL channel.

The invention discloses a method and device for detecting objects in a video and a computer storage medium. The method includes: determining a plurality of key frames according to the target video, and performing object detection on each key frame to obtain the detection result of each key frame; determining the detection result of an in-between frame between every two adjacent key frames; correcting the detection result of each in-between frame to obtain the corrected detection result of each in-between frame; determining the detection result of the target video according to the detection result of each key frame and the corrected detection result of each in-between frame.

A visual question and answer method based on a multi-modal decomposition model is provided. The method is characterized in that the image is trained in an ImageNet dataset and image features are extracted, the question is marked as texts into feature vectors; a collaborative attention model is introduced into the basic network architecture, image and question related features are learned, and fine-grained correlation among multi-modal features are characterized; and multi-modal features enter a multi-modal decomposition bilinear pool (MFB) or a multi-modal decomposition high order cell (MFH) module image to generate a fusion image question feature z, and the z is put into the classifier to predict the best matching answer. According to the method provided by the present invention, the collaborative attention model is used to predict the correlation between each spatial grid in the image and the question, and the best matching answer can be accurately predicted in a facilitated manner;and in combination with the image attention mechanism, the model can effectively understand which image region is important to the question, so that the performance of the model and the accuracy of the question and answer can be significantly improved.

A method for estimating noise in a reconstructed image through post-processing includes the steps: dividing the reconstructed image to generate image segments; applying a multi-resolution transformation or directional filter bank on at least part of the image segments to generate transformed image segments; and for each transformed image segment estimating a direction dependent noise power S₀(θ); calculating a first noise covariance matrix from an isotropic power spectral density |ω∥G(ω)|²; and calculating a second noise covariance matrix in the transformed image segment through the product of the direction dependent noise power S₀(θ) and the first noise covariance matrix.

A system and methods for weak spot analysis. A mesh shape is preprocessed to approximate an input model for an object. The mesh shape is analyzed with modal analysis to identify weak regions. A method and system for determining weak spots in an object. The method and system uses an optimization problem which is solved to determine a pressure distribution on the object maximizing maximal principal stress by solving a set of optimization problems maximizing stress for each point of the object.

The invention provide methods and apparatus for effectively identifying the occlusion of objects, such as persons, having a high degree of freedom. In an example embodiment, after initialization, an image is input, and an image region is extracted from image data. The distance is employed that is obtained when the shape of a two-dimensional histogram in the color space is transformed into the feature space. A graph is formed by using, the regions between the frames. A confidence factor is provided and image features are provided as weights to the edges that connect the nodes. Processing is performed, and the confidence factor is examined. A connection judged less possible to be a path is removed. When there is only one available connection for the occlusion point, this connection is selected.

The invention relates to a wind farm multi-model draught fan optimized arrangement method based on a genetic algorithm. The method includes the following steps that (1) a wind farm region is divided into square meshes which are the same in size according to the diameter of a draught fan, and an integer matrix which is the same in line and row is generated randomly to be used as the initial solution of the algorithm; (2) the individual fitness value of a current generation is calculated; (3) parent individuals participating in crossover are selected through even random selection operators, and then filial generation individuals are generated by the adoption of improved crossover and mutation operators; (4) repairing operators are introduced to the individuals in a population; (5) a Tabu operator is introduced to an optimal solution of the current generation of the population, the optimal solution is used as the initial solution of a Tabu algorithm, and the neighborhood solution of the optimal solution is searched for; (6) whether the biggest number of iterations is reached or not is judged, if yes, the multi-model draught fan optimized arrangement is completed, and if not, the step (2) is executed again. Compared with the prior art, the wind farm multi-model draught fan optimized arrangement method based on the genetic algorithm has the advantages of being visual in coding mode, good in performance index, high in local search capacity, high in expansibility, high in practicability and the like.

A technique is used to realize a generalized decision feedback equalizer (GDFE) Precoder for multi-user multiple-input multiple-output (MU-MIMO) systems, which significantly reduces the computational cost while resulting in no capacity loss. The technique is suitable for improving the performance of various MU-MIMO wireless systems including future 4G cellular networks. In one embodiment, a method for configuring a GDFE precoder in a base station of a MU-MIMO wireless system having k user terminals, each user terminal having associated therewith a feedforward filter. The method comprises computing a filter matrix C using one of a plurality of alternative formulas of the invention; and, based on the computation of the filter matrix C, computing a transmit filter matrix B for a transmit filter used to process a symbol vector obtained after a decision feedback equalizing stage of the GDFE precoder, a feedforward filter matrix F, and an interference pre-cancellation matrix G.

A method and apparatus for identifying a concrete crack includes: obtaining a crack video, and manually annotating a video image frame by using a label; predicting a future frame and label for the annotated frame by using a spatial displacement convolutional block, propagating the future frame and label, to obtain a synthetic sample, and preprocessing the synthetic sample, to form a crack database; modifying input and output ports of data of a deep learning model for video semantic image segmentation and a parameter, to enable the deep learning model to accept video input, and establishing a concrete crack detection model based on the video output; using a convolutional layer in a trained deep learning model as an initial weight of the concrete crack detection model for migration; inputting the crack database into a migrated concrete crack detection model, and training the concrete crack detection model for crack data.