2995results about How to "Reduce the amount of calculation" patented technology

One embodiment of the present invention determines the effect of placing an assist feature at a location in a layout. During operation, the system receives a first value which was pre-computed by convolving a model with a layout at an evaluation point, wherein the model models semiconductor manufacturing processes. Next, the system determines a second value by convolving the model with an assist feature, which is assumed to be located at a first location which is in proximity to the evaluation point. The system then determines the effect of placing an assist feature using the first value and the second value. An embodiment of the present invention can be used to determine a substantially optimal location for placing an assist feature in a layout.

A software simulation method and program storage device for software defect detection and obtaining insight into software code is disclosed, where simulation consists of executing target software program code for multiple input values and multiple code paths at the same time, thus achieving 100% coverage over inputs and paths without actually running the target software. This allows simulation to detect many defects that are missed by traditional testing tools. The simulation method runs a plurality of algorithms where a plurality of custom defined and pre-defined rules are verified in target software to find defects and obtain properties of the software code.

An object detector, an object detecting method and a robot can reduce diction errors of detecting wrong objects without increasing the volume of the computational operation to be performed to detect the right object. A face detector 101 comprises a face detecting section 110 that operates like a conventional face detecting section and is adapted to roughly select face candidates from an input image by template matching and detect face candidates by means of a support vector machine for face recognition, a non-face judging section 120 that detects non-face candidates that are judged to be non-faces and removes them from the face candidates selected by the face detecting section 110 and a skin tracker section 114 for tracking a face region after the non-face judgment. When the assumed distance between the face detector and the face as computed from the input image and the measured distance as measured by a distance sensor show a large difference, when the color variance of the face candidate is small, when the occupancy ratio of the skin color region is large and when the change in the size of the face region is large after the elapse of a predetermined time, the non-face judging section 120 judges such face candidates as non-faces and removes them from the face candidates.

A consistent tie-breaking decision between equal-cost shortest (lowest cost) paths is achieved by comparing an ordered set of node identifiers for each of a plurality of end-to-end paths. Alternatively, the same results can be achieved, on-the-fly, as a shortest path tree is constructed, by making a selection of an equal-cost path using the node identifiers of the diverging branches of the tree. Both variants allow a consistent selection to be made of equal-cost paths, regardless of where in the network the shortest paths are calculated. This ensures that traffic flow between any two nodes, in both the forward and reverse directions, will always follow the same path through the network.

A mobile robot having a movable main unit section, a self location measurement unit for measuring a self location of the main unit section, a map database for storing map information on a travel range of the main unit section to a travel destination, a virtual sensor information calculation unit for extracting information on obstacles to movement of the main unit section in an arbitrary detection region on the map information based on self location information measured by the self location measurement unit and the map information stored in the map database, and for calculating virtual sensor calculation information, and a route calculation unit for calculating a travel route for the main unit section to travel based on the virtual sensor calculation information calculated by the virtual sensor information calculation unit.

It is proposed a text-independent automatic speaker recognition (ASkR) system which employs a new speech feature and a new classifier. The statistical feature pH is a vector of Hurst parameters obtained by applying a wavelet-based multi-dimensional estimator (M dim wavelets) to the windowed short-time segments of speech. The proposed classifier for the speaker identification and verification tasks is based on the multi-dimensional fBm (fractional Brownian motion) model, denoted by M dim fBm. For a given sequence of input speech features, the speaker model is obtained from the sequence of vectors of H parameters, means and variances of these features.

An energy corrector (105) for correcting a target energy for high-frequency components and a corrective coefficient calculator (106) for calculating an energy corrective coefficient from low-frequency subband signals are newly provided. These processors perform a process for correcting a target energy that is required when a band expanding process is performed on a real number only. Thus, a real subband combining filter and a real band expander which require a smaller amount of calculations can be used instead of a complex subband combining filter and a complex band expander, while maintaining a high sound-quality level, and the required amount of calculations and the apparatus scale can be reduced.

A consistent tie-breaking decision between equal-cost shortest (lowest cost) paths is achieved by comparing an ordered set of node identifiers for each of a plurality of end-to-end paths. Alternatively, the same results can be achieved, on-the-fly, as a shortest path tree is constructed, by making a selection of an equal-cost path using the node identifiers of the diverging branches of the tree. Both variants allow a consistent selection to be made of equal-cost paths, regardless of where in the network the shortest paths are calculated. This ensures that traffic flow between any two nodes, in both the forward and reverse directions, will always follow the same path through the network.

A digital imaging system is described that provides techniques for reducing the amount of processing power required by a given digital camera device and for reducing the bandwidth required for transmitting image information to a target platform. The system defers and / or distributes the processing between the digital imager (i.e., digital camera itself) and the target platform that the digital imager will ultimately be connected to. In this manner, the system is able to decrease the actual computation that occurs at the digital imager. Instead, the system only performs a partial computation at the digital imager device and completes the computation somewhere else, such as at a target computing device (e.g., desktop computer) where time and size are not an issue (relative to the imager). By deferring resource-intensive computations, the present invention substantially reduces the processor requirements and concomitant battery requirements for digital cameras. Further, by adopting an image strategy optimized for compression (compressed luminosity record), the present invention decreases the bandwidth requirements for transmitting images, thereby facilitating the wireless transmission of digital camera images.

An apparatus and a method for peak-to-average power ratio reduction of an OFDM signal are disclosed. The method uses the interleaved characteristics of partial transmit sequences to partition input data x[n] of length N into several disjoint subblocks in time domain, and a complete N-point signal {tilde over (x)}[n] is composed after phase optimization, where N is the length of an OFDM signal and n=0, 1, . . . , N−1. Accordingly, the apparatus comprises an N-point inverse fast Fourier transform (N-IFFT), a de-multiplexer, a combiner, a set of memory and an adder. This invention uses only one N-IFFT, whereby it can achieve significant computation reduction. This invention requires (N / 2)log2 N complex multiplications and N memory units. It also preserves the inherent property as well as advantages of an OFDM system.

The invention discloses a multi-scale image semantic segmentation method. The method comprises the following steps: obtaining a to-be-segmented image and a corresponding label; constructing a full convolutional deep neural network, wherein the full convolutional deep neural network comprises a convolution module, a hole convolution module, a pyramid pooling module, a 1 * 1 * depth convolution layer and a deconvolution structure; setting hole convolution as channel-by-channel operation, and utilizing low-scale, medium-scale and high-scale characteristics in a targeted mode; training the full convolutional deep neural network, establishing a loss function, and determining parameters of the full convolutional deep neural network by training the sample image; and inputting the to-be-segmentedimage into the trained full convolutional deep neural network to obtain a semantic segmentation result. By means of the method, the image semantic segmentation problem with complex details, holes andlarge targets can be well solved while the calculated amount and the parameter number are reduced, and the consistency of category labels can be reserved while the target edges can be well segmented.

An apparatus and method for cell acquisition and downlink synchronization acquisition in an OFDMA wireless communication system are provided. In an SS apparatus in a broadband wireless communication system, a preamble subcarrier acquirer extracts subcarrier values having a preamble code from an FFT signal. A multiplier code-demodulates the subcarrier values by multiplying the subcarrier values by a preamble code. A correlator calculates a plurality of differential correlations in the code-demodulated signal. An IFFT processor IFFT-processes the differential correlations by mapping the differential correlations to subcarriers. A maximum value detector detects a maximum value from the IFFT signal and calculates a timing offset using an IFFT output index having the maximum value.

A method of computing a manufacturing yield of an integrated circuit having device shapes includes sub-dividing the integrated circuit into failure mechanism subdivisions (each of the failure mechanism subdivisions includes one or more failure mechanism and each of the failure mechanisms includes one or more defect mechanisms), partitioning the failure mechanism subdivisions by area into partitions, pre-processing the device shapes in each partition, computing an initial average number of faults for each of the failure mechanisms and for each partition by numerical integration of an average probability of failure of each failure mechanism, (the numerical integration produces a list of defect sizes for each defect mechanism, and the computing of the initial average includes setting a maximum integration error limit, a maximum sample size for a population of each defect size, and a maximum number of allowable faults for each failure mechansim), and computing a final average number of faults for the integrated circuit by iterativelly reducing a statistical error of the initial average number of faults for each of the failure mechanisms until the statistical error is below an error limit.

The invention discloses a face recognition method and system, in particular relates to a face recognition model training method and system. The face recognition method comprises the following steps: every user for recognition is trained and a face recognition model is obtained; within the set face posture range, the face image with structure light coding and the face image without structure light coding of the present user waiting for recognition are collected simultaneously; a face range image is obtained based on the face image with structure light coding, and a face gray image is obtained based on the face image without structure light coding; the face gray image is used for calibration, and face pre-treatment is conducted for the face gray image and the face range image based on the calibration results; the characteristics of pretreated face range image and face gray image are calculated respectively; the characteristics are input to the face recognition model to recognize faces and output face recognition results. The technical proposal provided by the invention can reduce the interference of light to face recognition and reduce costs simultaneously.

The invention provides a vehicle license plate recognition method based on a video. According to the vehicle license plate recognition method based on the video, moving vehicles are detected and separated out with the vehicle video which is obtained through actual photographing by means of a camera serving as input, the accurate position of a vehicle license plate area is determined by conducting vertical edge extraction on a target vehicle image obtained after pre-processing, a vehicle license plate image is separated out, color correction, binaryzation and inclination correction are conducted on a vehicle license plate image, each character in the positioned vehicle license plate area is separated to serve as an independent character, feature extraction is conducted one each character, obtained feature vectors are classified through a classifier which is well trained in advance, a classification result serves as a preliminary recognition result, secondary recognition is conducted on the stained vehicle license plate characters according to a template matching algorithm imitating the visual characteristics of human eyes, and then a final vehicle license plate recognition result is obtained. The vehicle license plate recognition method based on the video has the advantages that hardware cost is reduced, the management efficiency of an intelligent transportation system is improved, the anti-jamming performance and the robustness are high, the recognition efficiency is high, and the recognition speed is high.

The invention provides a method and a system for transactions between a main chain and a side chain of a block chain. According to the invention, through the side chain, service types of the block chain can be expanded; by use of a flow dividing mode, transaction quantity of the main chain is reduced; and nodes of the main chain and nodes of the side chain store and verify block data and transaction data of own chains respectively. Thus, data quantity and calculated amount of the transaction are reduced and problems in data storage and expansion of the calculated amount of a single node caused by the fact that different services are supported by the same chain are solved. In addition, by use of a de-centralization scheme, bidirectional anchoring of capitals between the side chain and the main chain can be achieved; multiple side chains are capable of using or anchoring capitals on the same main chain; capital types are reduced; a mutual exchange problem between different capitals is avoided; and the services are quite clear and rational.

The invention discloses a method for self-adaptive subsection interpolation of NURBS curve under the real-time look-ahead whole-process accelerated and decelerated control. The NURBS curve is presented by a matrix form and is pre-processed; the self-adaptive speed adjusting arithmetic is adopted for the NURBS curve, thus obtaining the parameters meeting the interpolation precision requirement such as feeding speed and the like; the curve is segmented at all maximum curvatures of the curve; speed planning is carried out on all the segmented curves by an S-curve accelerated and decelerated control method, thus obtaining the changing time of the speed curve at all accelerated and decelerated stage; according to the interpolation period and the changing time of all accelerated and decelerated stage, the feeding speed of all interpolation periods and coordinates of corresponding interpolation points are obtained; and according to the coordinates of the interpolation points and the feeding speed, the control signal given value is generated. The invention can realize the look-ahead processing of the NURBS curve during the whole interpolation process, leads the real-time interpolation process to be quick and effective, leads the whole interpolation movement process to be smooth and stable, has wide applicable range and has important significance to the development of the high-speed high-precision numerical control system.

A method of determining an initial position in executing a positioning calculation based on satellite signals from positioning satellites, includes: (a) executing position converging calculation based on the satellite signals using each of a plurality of first candidate positions as a provisional initial position; (b) selecting at least one second candidate position from the first candidate positions based on a calculation result of the position converging calculation; and (c) selecting the initial position from one of the second candidate positions using the satellite signals.

To provide a bandwidth extension method which allows reduction of computation amount in bandwidth extension and suppression of deterioration of quality in the bandwidth to be extended. In the bandwidth extension method: a low frequency bandwidth signal is transformed into a QMF domain to generate a first low frequency QMF spectrum; pitch-shifted signals are generated by applying different shifting factors on the low frequency bandwidth signal; a high frequency QMF spectrum is generated by time-stretching the pitch-shifted signals in the QMF domain; the high frequency QMF spectrum is modified; and the modified high frequency QMF spectrum is combined with the first low frequency QMF spectrum.

The present invention relates to an image processing apparatus and a method, and in particular to an image processing apparatus and a method preferably applicable to conversion of a wide dynamic-range image having a dynamic range of pixel values wider than the normal one to a narrow dynamic-range image having a narrower dynamic range of pixel values, and to enhancement of contrast. In step S1, an input wide-DR luminance image of the current frame is converted into a narrow-DR luminance image based on the intermediate information calculated for the previous frame's wide-DR luminance image. In step S2, the stored intermediate information of the previous frame is updated using the calculated intermediate information. In step S3, it is determined if there is any succeeding frame. If there is the succeeding frame, the process returns to step S1 and processes thereafter are repeated. The present invention is applicable to a digital video camera and the like.

A training circuit for training a tap coefficient of an adaptive equalizer that performs error calculations in a training process and updating of the tap coefficient in a time area and thereby makes the tap coefficient converge stably in a short time.

The invention discloses a depth learning target detection method based on pre-segmentation and regression, mainly to solve the problems that the existing target detection method is poor in small target detection precision and long in detection time. The method comprises steps: 1) a quadtree segmentation algorithm is used to extract a region of interest in a to-be-detected image; 2) a basic convolution layer and an auxiliary convolution layer are used to carry out feature extraction on the region of interest, and feature graphs of multiple scales are obtained; 3) the position information of a default border is calculated on the feature graphs of multiple scales, a convolution filter is used for detection on the feature graphs of multiple scales, and multiple predicted borders and multiple category scores are obtained; and 4) non-maximum suppression is used for the multiple predicted borders and the multiple category scores, and the final target border position and the category information are obtained. A small target in the image can be quickly and accurately detected, and the method can be used for target real-time detection in unmanned aerial vehicle aerial photographing.

System and method of estimating radio channel bit error rate (BER) in a digital radio telecommunications system wherein the soft output of the turbo decoder is used as pointer or index to look-up-tables containing the bit-wise BER of a certain bit in the data field of the received frame. A quantizer quantizes the received data frame and the quantized bit operates on a switch which selects the appropriate look-up-table. By means of accumulation and scaling the average BER of a certain amount of bits are calculated. Decoding bit-errors may occur but as they are submitted to posterior probability estimation, systematic errors which normally happen at low SNR are avoided.

This invention deals with an advanced Real-time Grid Monitoring System (RTGMS) suitable for both single-phase and three-phase electric power systems. This invention provides an essential signal processing block to be used as a part of complex systems either focused on supervising and diagnosing power systems or devoted to control power processors interacting with the grid. This invention is based on a new algorithm very suitable for real-time characterization of the grid variables under distorted and unbalanced grid conditions. The main characteristic of this invention is the usage of a frequency-locked loop, based on detecting the grid frequency, for synchronizing to the grid variables. It results in a very robust system response in relation to existing technique based on the phase-angle detection since grid frequency is much more stable variable than the grid voltage / current phase-angle, mainly during grid faults. Moreover, the algorithm supporting this invention is very efficient and can be implemented in regular industrial microprocessors. These features make the RTGMS object of this invention ideal to be applied in the control of distributed generation systems (DGS), flexible AC transmission systems (FACTS), power quality conditioners (PQC) and uninterruptible power supplies (UPS). In all these systems, the fast and precise real time detection of the voltage and / or current sequence components under grid fault conditions is a crucial matter.

Disclosed herein are a markerless augmented reality system and method for extracting feature points within an image and providing augmented reality using a projective invariant of the feature points. The feature points are tracked in two images photographed while varying the position of an image acquisition unit, a set of feature points satisfying a plane projective invariant is obtained from the feature points, and augmented reality is provided based on the set of feature points. Accordingly, since the set of feature points satisfies the plane projective invariant even when the image acquisition unit is moved and functions as a marker, a separate marker is unnecessary. In addition, since augmented reality is provided based on the set of feature points, a total computation amount is decreased and augmented reality is more efficiently provided.

A receiver and transmitter of a closed-loop MIMO antenna system using a codebook and a receiving and transmitting method thereof are provided. The receiver of the MIMO antenna system includes a window size decider and a beamforming weight selector. The window size decider stores a codebook with beamforming weights and selects the beamforming weights corresponding to a window size from the codebook, and the beamforming weight selector selects an optimal beamforming weight based on a current channel state among the beamforming weights outputted from the window size decider, and feeds back the selected optimal beamforming weight to a transmitter.

A method of evaluating the quality of a read signal from the viewpoint of the detection margin of a Viterbi decoder in the PRML method in which a target signal level varies depending on the read signal, and an optical disc apparatus implementing the method. A method of evaluating the quality of a signal pattern comprising a combination of minimum run lengths from the viewpoint of edge shift, and an optical disc apparatus implementing the method. From the target signal level that varies depending on the read signal, a target signal is generated based on a decoding result, and an error target signal is generated in which the decoding result is edge-shifted. The signal quality is evaluated by calculating a Euclidean distance between these signals and the read signal. A virtual state that is not included in the Viterbi decoder and that is less than the minimum run length is defined, and a target signal level for the virtual state is generated using a target signal level table inside the Viterbi decoder, based on the concept of convolution. In this way, the signal quality can be evaluated by the same method as mentioned above even in cases where the pattern of a combination of the minimum run lengths has edge-shifted.

The invention discloses a non-overlapping field-of-view camera gesture calibration method based on point cloud feature map registration. The method comprises the following steps that: (1) carrying outbasic calibration on a plurality of cameras of a non-overlapping field of view to obtain an internal reference; (2) utilizing the plurality of cameras to carry out environment detection and synchronous positioning and mapping, constructing a point cloud map, and extracting a key frame to solve the pose matrix of the camera; (3) extracting an image frame from the key frame of one camera, carryingout similarity detection on the key frames of other cameras, constructing a matching frame point set and a matching point pair set, and carrying out projection error minimization on the projection ofthe point cloud map point on the image frame and the practical pixel coordinate; and (4) carrying out feature matching on a frame near the matched frame, blending all feature points, carrying out global optimization and iterative solution on a relative pose matrix, selecting a correction parameter according to a practical situation, and carrying out final gesture calibration on the camera. By useof the method, the problems of high calibration work intensity, low work efficiency and low accuracy of a traditional calibration method are solved.

The invention discloses an image appearance based loop closure detecting method in monocular vision SLAM (simultaneous localization and mapping). The image appearance based loop closure detecting method includes acquiring images of the current scene by a monocular camera carried by a mobile robot during advancing, and extracting characteristics of bag of visual words of the images of the current scene; preprocessing the images by details of measuring similarities of the images according to inner products of image weight vectors and rejecting the current image highly similar to a previous history image; updating posterior probability in a loop closure hypothetical state by a Bayesian filter process to carry out loop closure detection so as to judge whether the current image is subjected to loop closure or not; and verifying loop closure detection results obtained in the previous step by an image reverse retrieval process. Further, in a process of establishing a visual dictionary, the quantity of clustering categories is regulated dynamically according to TSC (tightness and separation criterion) values which serve as an evaluation criterion for clustering results. Compared with the prior art, the loop closure detecting method has the advantages of high instantaneity and detection precision.