Abstract: A method is provided. The method includes receiving a first dimension set, extracting a first latent feature set from the first dimension set, training a first base predictor based on the first feature set, generating a second dimension set based on the first dimension set, the second dimension set having fewer dimensions than the first dimension set, extracting a second latent feature set from the second dimension set, training a second base predictor based on the second feature set, and generating a traffic prediction based on the first base predictor and the second base predictor.

Abstract: A method of training a neural network based three-dimensional (3D) encoder is provided. A training dataset is generated using a plurality of 3D models of a 3D model dataset. To generate a first sample of the training dataset, an image generator with multi-view rendering is used to generate a plurality of image candidates of a first 3D model. A word is chosen from metadata associated with the first 3D model. A language model is used to generate one or more text descriptions using the selected word and a plurality of prompts. A point cloud is generated by randomly sampling points in the 3D model. The first sample is generated to include a first image randomly selected from the plurality of image candidates, one or more text descriptions, and the point cloud is generated. The 3D encoder is trained using the training dataset including the first sample.

Abstract: The present disclosure provides a display substrate and a display device, and belongs to the field of display technologies.

Abstract: The present invention discloses a microminiature image acquisition and processing system package structure and a preparation method thereof. This structure includes optical coated glass, a CMOS chip, a wafer Re-Distribution Layer and a molding layer, the first surface of the CMOS chip is provided with a photosensitive and microlens region and a metal bonding pad, and a through-silicon via is etched in a second surface of the CMOS chip until it extends to the metal bonding pad on the first surface; the wafer Re-Distribution Layer covers the second surface of the CMOS chip and extends to the through-silicon via. The structure and the method of the present invention are integrated with wafer-level package and SIP integrated package technologies to achieve single package of the whole device, thereby greatly reducing the system complexity and power consumption, reducing the overall product size and signal path, and improving the image anti-interference capability.

Abstract: A driving circuit and a driving method are provided. The driving circuit includes a charging circuit and a discharging circuit. The charging circuit is configured to receive an input voltage to charge the piezoelectric load. The piezoelectric load discharges electricity through the discharging circuit. Operation states of the charging circuit and the discharging circuit are controlled. During a first operation interval of an operation period, the charging circuit charges the piezoelectric load so that a power supply voltage signal provided to the piezoelectric load corresponds to the reference voltage in a first interval. During a second operation interval of the operation period, the piezoelectric load discharges electricity through the discharging circuit so that the power supply voltage signal corresponds to the reference voltage in a second interval. The driving circuit according to the present disclosure requires a few switches, thereby facilitating circuit integration.

Abstract: The present disclosure provides a display substrate, in which power lines and compensation detection lines are alternately arranged at intervals, and two columns of pixel driving circuits extending along a second direction are arranged between any one of the power lines and one adjacent compensation detection line; and for any one of a plurality of pixel driving circuits, a power input terminal of the pixel driving circuit is electrically connected to the power line closest to the pixel driving circuit, and a compensation detection signal terminal of the pixel driving circuit is electrically connected to the compensation detection line closest to the pixel driving circuit. The present disclosure further provides a display device.

Abstract: The disclosure provides a display panel and a display device. The display panel includes: a base substrate, a transistor array layer, a pixel defining layer, touch electrodes. The area of opening region of first color sub-pixel is smaller than that of opening region of third color sub-pixel, the area of opening region of second color sub-pixel is smaller than that of opening region of third color sub-pixel. An orthogonal projection of second capacitor in first color sub-pixel and an orthogonal projection of touch electrodes have a first auxiliary overlap area, an orthogonal projection of second capacitor in second color sub-pixel and orthogonal projection of the touch electrodes have a second auxiliary overlap area, an orthogonal projection of second capacitor in third color sub-pixel and orthogonal projection of touch electrodes have a third auxiliary overlap area. The first and/or second auxiliary overlap area is larger than the third auxiliary overlap area.

Abstract: A model training method and a related apparatus to help improve a convergence speed of model training and improve end-to-end communication quality. The method includes: a first communication apparatus sends first data to a second communication apparatus through a channel, where the first data is an output result of the first machine learning model. The second communication apparatus receives second data through a channel, inputs the second data into a second machine learning model to obtain third data; determines a first loss function based on the third data and the first training data; and sends the first loss function to the first communication apparatus through a feedback channel.

Abstract: A method, system, and article is directed to automatic content-dependent image processing algorithm selection.

Abstract: The present invention provides a virtual keyboard for inputting Chinese characters and a configuring method thereof, an input method, and a Chinese character input system. The configuring method includes the following steps: setting a geometric layout of the virtual keyboard; and setting initials and finals on available keys of the virtual keyboard respectively using a cost function-based search algorithm so as to obtain an optimal layout of initial keys and final keys, wherein the initial keys and the final keys in the optimal layout are set based on a minimum cost function value for spelling all Chinese syllables, and the minimum cost function value is a minimum sum of weighted distances of the initial keys and the final keys for all the Chinese syllables.

Abstract: The present invention provides a dual-substrate antenna package structure and a method for manufacturing the same. The package structure includes a main substrate and at least one antenna substrate. The antenna substrate is provided on a pad of the main substrate by an array of solder balls placed on the antenna substrate, at least one chip is electrically connected to the main substrate, and metal wiring provided on the main substrate electrically connects the pad to the chip. The array of solder balls includes support solder balls and conventional solder balls, and the support solder balls have a melting point high than 250° C. A spacing distance between the antenna substrate and the main substrate can be kept stable during the reflow soldering process and subsequent processes because the support solder balls having the high melting point can always maintain the stability of the structure during the reflow soldering process.

Abstract: A display substrate is provided. The display substrate includes a base substrate including a display region and a peripheral region, and a first scan driving circuit, a second scan driving circuit and a first power line arranged in sequence, and a first shielding layer and a second shielding layer sequentially arranged on a side of the second scan driving circuit away from the base substrate; the first shielding layer covers at least one transistor in the second scan driving circuit, and the second shielding layer covers at least one transistor of transistors in the second scan driving circuit except the at least one transistor covered by the first shielding layer; and the second shielding layer is also on a side of the first scan driving circuit away from the base substrate, and the second shielding layer covers at least one transistor in the first scan driving circuit.

Abstract: This application relates to the field of wireless communications technologies, and discloses an encoding method and apparatus, to improve accuracy of reliability calculation and ordering for polarized channels. The method includes: obtaining a first sequence used to encode K to-be-encoded bits, where the first sequence includes sequence numbers of N polarized channels, the first sequence is same as a second sequence or a subset of the second sequence, the second sequence comprises sequence numbers of Nmax polarized channels, and the second sequence is the sequence shown in Sequence Q11 or Table Q11, K is a positive integer, N is a positive integer power of 2, n is equal to or greater than 5, K?N, Nmax=1024; selecting sequence numbers of K polarized channels from the first sequence; and performing polar code encoding on K the to-be-encoded bits based on the selected sequence numbers of the K polarized channels.

Abstract: A display device and a manufacturing method thereof are disclosed. The display device includes a base substrate and at least one pixel circuit provided on the base substrate, and the pixel circuit includes a driving transistor, a first transistor, and a second transistor; the driving transistor includes a control electrode, a first electrode, and a second electrode; the first transistor includes a first active region, the second transistor includes a second active region, the driving transistor includes a fourth active region, at least one of a doping concentration of the first active region and a doping concentration of the second active region is greater than a doping concentration of the fourth active region.

Abstract: An inspection system includes a ground penetrating radar (GPR) sensor configured to provide at least one probe signal to a portion of a structure and to receive at least one return signal resulting from the corresponding at least one probe signal. A processor is programmed to: receive the at least one return signal as scan data from the GPR sensor; generate maximum amplitude trace normalized data based on the scan data; generate temporal signal gain data based on the scan data; generate a power spectral density image for each of the scan data, the maximum amplitude trace normalized data, and the temporal signal gain data; and analyze, using a machine learning processor, the power spectral density images to identify an anomaly of interest, wherein the machine learning processor is trained to detect an anomaly of interest based on a training set of power spectral density images.

Abstract: A detection device includes a frame, a transport mechanism, detection mechanisms, and a grasping mechanism. The transport mechanism includes a feeding line, a first flow line, and a second flow line arranged in parallel on the frame. The detection mechanisms are arranged on the frame and located on two sides of the transport mechanism. The grasping mechanism is arranged on the frame and used to transport workpieces on the feeding line to the detection mechanisms, transport qualified workpieces to the first flow line, and transport unqualified workpieces to the second flow line.

Abstract: To automate time series forecasting machine learning pipeline generation, a data allocation size of time series data may be determined based on one or more characteristics of a time series data set. The time series data may be allocated for use by candidate machine learning pipelines based on the data allocation size. Features for the time series data may be determined and cached by the candidate machine learning pipelines. Predictions of each of the candidate machine learning pipelines using at least the one or more features may be evaluated. A ranked list of machine learning pipelines may be automatically generated from the candidate machine learning pipelines for time series forecasting based upon evaluating predictions of each of the one or more candidate machine learning pipelines.

Abstract: A wearable device includes a frame element and a dielectric substrate. The frame element includes a first metal element, a second metal element, and a third metal element. A first gap is provided between the first metal element and the second metal element. A second gap is provided between the second metal element and the third metal element. A third gap is provided between the third metal element and the first metal element. The dielectric substrate is surrounded by the first metal element, the second metal element, and the third metal element. A first antenna element is formed by the first metal element. A second antenna element is formed by the second metal element. A third antenna element is formed by the third metal element.

Abstract: A method of monitoring building compliance with healthy building guidelines includes obtaining current parameter values for a plurality of different parameters from a plurality of sensors disposed within a plurality of different zones of a building. For each of the parameters, the current parameter value received from each of the plurality of different zones is compared with a corresponding healthy building range for that parameter as specified by the healthy building guidelines. A healthy building dashboard is displayed that includes a summary that shows, for each of the plurality of different parameters, how many zones of the plurality of different zones of the building are within the corresponding healthy building range for that parameter and/or how many zones of the plurality of different zones are not within the corresponding healthy building range for that parameter.

Abstract: There is provided a display substrate and a display device. The display substrate includes a first metal layer, a first insulating layer, a metal oxide layer, a second insulating layer and a second metal layer which are stacked; wherein the metal oxide layer comprises a first pattern, a second pattern and a capacitance pattern, the first metal layer comprises a first electrode plate, there is at least a first overlapping region between the first electrode plate and the capacitance pattern to form a first storage capacitor, the second metal layer comprises a second electrode plate, there is at least a second overlapping region between the second electrode plate on the base substrate and the capacitance pattern to form a second storage capacitor, and the first electrode plate and the second electrode plate have same potential.