54results about How to "Reduce estimation bias" patented technology

A neural network interlayer activation value quantification method and apparatus, wherein the method comprises: acquiring an activation value of the previous activation layer of a neural network; obtaining an activation value limit of a current activation layer according to the activation value of the previous activation layer; determining respective activation values of the current activation layer according to the activation value limit of the current activation layer and the activation value of the previous activation layer. Therefore, the activation value information of the previous activation layer can be transmitted to the current activation layer. Compared with the fixedly set limit in the prior art, the technical solution provided by the present invention can dynamically adapt to the change of the activation value according to the activation value of the previous activation layer, thereby reducing the estimated deviation of quantization.

The invention provides a signal phase difference estimation device in an antenna aligning array based on circulating cross-correlation, and the antenna aligning array comprises a plurality of receiving antenna units. The device is used for calculating the phase differences of antenna signals of the antenna aligning array in stable circulation, and comprises one or two matched filters, a circulating cross-correlation module and a phase difference resolving module, wherein the filter is used for matched filtering for the antenna signal which is subjected to time delay, amplitude weighting and phase adjustment, and matched filtering for a difference signal between the signal which is subjected to time delay, amplitude weighting and phase adjustment and a reference signal; the circulating cross-correlation module is used for calculating a circulating cross-correlation function of the signal and the difference signal which are both subjected to matched filtering; and the phase difference resolving module is used for calculating the phase difference of the antenna signal according to the circulating cross-correlation function. The invention also provides a signal phase difference estimating method in the antenna aligning array based on circulating cross-correlation, a signal synthesis device, a signal synthesizing method, an antenna aligning array signal receiving system and an antenna aligning array signal receiving method.

The invention discloses a non-data auxiliary SNR estimation method for open loop and closed loop combined MPSK signals. The method includes the first step of obtaining second-order envelope statistical information and first-order envelope statistical information through two received orthogonal signals, the second step of utilizing the square of the sum of the second-order statistical information of first-section data and the square of the sum of the first-order statistical information of the first-section data to calculate a statistical parameter value, the third step of estimating the SNR according to the one-one mapping relation between the statistical parameter value and the SNR, and the fourth step of carrying out smoothing filtering on the SNR estimated value through a closed-loop filter. The method has the advantages that any auxiliary data are not needed, carrier recovery does not need to be carried out, the SNR estimated value can be fast obtained, the estimation accuracy is high, the estimation deviation is small, the requirements of data quantization in the hardware implementation process are fully considered, and the complexity of hardware implementation is low. The non-data auxiliary SNR estimation method for the open loop and closed loop combined MPSK signals is applied to a communication system using an MPSK constant envelop modulation mode and other modes.

The present application provides an image processing method, an image processing device, and a terminal device. The method includes: acquiring multiple images, and judging whether the multiple images are images taken in the same area; if the multiple images are For images taken in the same area, determine the overlapping background areas of the multiple images; count the color histogram of the overlapping background area; judge whether the overlapping background area is a preset solid color background according to the color histogram ; If the overlapping background area is a preset solid-color background, select an image from the plurality of images, and perform white balance processing on the selected image according to the overlapping background area. The present application can solve the problem that the color cast of the processed image caused by the traditional white balance algorithm is too serious when the image has a large solid color background.

The invention discloses a piecewise constant independent component analysis-based method for estimating a harmonic current of a power distribution network. The method comprises the steps of firstly building a piecewise constant model of harmonic current estimation, determining a harmonic current sequence by adopting a correlation coefficient matrix association method of overlapped sections, determining a harmonic current amplitude by utilizing an error minimization method of the overlapped sections, and lowering the need of prior harmonic current data; secondly giving a criterion of power network parameter change; and finally determining a data interval of the power network parameter change through a bisection method, and correcting a harmonic current estimation result during the power network parameter change to improve estimation precision during the power network parameter change. According to the method, harmonic current estimation errors during the power network parameter change can be reduced and the estimation process is free of power network parameter data; and the method is suitable for the harmonic current estimation under the condition that the parameters of the power distribution network are dynamically changed.