Abstract: A method for preparing a rubber for an elastomer of a flexible joint, includes: weighing hydrogenated nitrile raw rubber, a silica reinforcing agent, aramid fiber pulp, carbon black, liquid nitrile rubber, zinc oxide, stearic acid, an antioxidant, and a peroxide crosslinking agent; plasticating the hydrogenated nitrile raw rubber in a first internal mixer at 125-135° C., collecting a resulting product and cooling, to yield a plasticized rubber; preheating the liquid nitrile rubber; adding the plasticized rubber to a second internal mixer, further adding the aramid fiber pulp, the zinc oxide, the stearic acid, and the antioxidant to the second internal mixer for first mixing; adding the carbon black to the second internal mixer for second mixing; adding the silica reinforcing agent and the preheated liquid nitrile rubber to the second internal mixer for third mixing; and collecting a discharged product when a temperature of the second internal mixer is 140° C.

Abstract: The architectural complexity of a neural network is reduced by selectively pruning channels. A cost metric for a convolution layer is determined. The cost metric indicates a resource cost per channel for the channels of the layer. Training the neural network includes, for channels of the layer, updating a channel-scaling coefficient based on the cost metric. The channel-scaling coefficient linearly scales the output of the channel. A constant channel is identified based on the channel-scaling coefficients. The neural network is updated by pruning the constant channel. Model weights are updated via a stochastic gradient descent of a training loss function evaluated on training data. The channel-scaling coefficients are updated via an iterative-thresholding algorithm that penalizes a batch normalization loss function based on the cost metric for the layer and a norm of the channel-scaling coefficients.

Abstract: The present disclosure provides a display panel having a display area and including: a plurality of infrared emitting units and a plurality of infrared receiving units in the display area. The plurality of infrared emitting units are configured to emit infrared light to outside of the display panel, and the plurality of infrared receiving units are configured to receive the infrared light reflected by an object outside the display panel and output an electrical signal according to the received infrared light.

Abstract: The disclosure discloses a battery, an electronic device and a battery charging method. When it is determined that a charging strategy needs to be updated, the charge cut-off voltage and charging rate of each charging stage in the current charging process may be determined, so that the battery is charged based on the determined charge cut-off voltage and charging rate, thereby adjusting and updating the charging strategy.

Abstract: A method of improving quality of crowdsourced affective data based on agreement relationship between a plurality of annotators include receiving, by a processor, a collection of stimuli previously given affective labels by the plurality of annotators, executing, by a processor, an algorithm operative to perform the steps including constructing an agreement multigraph as a probabilistic model including a pair-wise status of agreement between the affective labels given by different ones of the plurality of annotators, learning the probabilistic model computationally using the crowdsourced affective data, identifying a reliability of each of the plurality of annotators based on the learned model, and adjusting the crowdsourced affective data by calculating the affective labels of each stimuli based on the identified reliability of each of the plurality of annotators, thereby improving the quality of the crowdsourced affective data.

Abstract: A silicon-graphite composite, a preparation method thereof, and a lithium battery anode and a lithium battery containing the silicon-graphite composite are provided in an embodiment of the disclosure. The silicon-graphite composite includes graphite and a silicon source fiber. The silicon source fiber is embedded in an interlayer structure of the graphite. The silicon-graphite composite is used as an anode material of the lithium battery in an embodiment of the disclosure.

Abstract: The present disclosure provides a display panel having a display area and including: a plurality of infrared emitting units and a plurality of infrared receiving units in the display area. The plurality of infrared emitting units are configured to emit infrared light to outside of the display panel, and the plurality of infrared receiving units are configured to receive the infrared light reflected by an object outside the display panel and output an electrical signal according to the received infrared light.

Abstract: Disclosed are an alignment mask, a metal mask assembly, and a preparation method therefor. The alignment mask includes a mask body. The mask body includes: multiple alignment holes; and, separating parts surrounding at least some alignment holes and used for separating the areas at where the at least some alignment holes are located from other areas, where the separation parts include at least one semi-etched line.

Abstract: A silicon-graphite composite, a preparation method thereof, and a lithium battery anode and a lithium battery containing the silicon-graphite composite are provided in an embodiment of the disclosure. The silicon-graphite composite includes graphite and a silicon source fiber. The silicon source fiber is embedded in an interlayer structure of the graphite. The silicon-graphite composite is used as an anode material of the lithium battery in an embodiment of the disclosure.

Abstract: A touchscreen includes a touch drive electrode, a touch sensing electrode, a touch drive circuit, a gate drive circuit, a touch drive electrode lead, and a liquid crystal panel. The touch drive electrode extends along a first direction. The touch drive circuit is disposed in a non-display area of the touchscreen, and an extension direction of the touch drive circuit is parallel to an extension direction of the touch drive electrode. The extension direction of the touch drive circuit is the same as the extension direction of the touch drive electrode. Each touch drive electrode is connected to the touch drive circuit using a touch drive electrode lead. The touch drive electrode lead is disposed in an area in which a gate drive circuit is located. The touch drive electrode lead and the gate drive circuit are disposed in an overlapping manner and are isolated from each other.

Abstract: An emotion analysis and recognition system including an automated recognition of bodily expression of emotion (ARBEE) system is described. The system may include program instructions executable by a processor to: receive a plurality of body movement models, each body movement model generated based on a crowdsourced body language dataset, calculate at least one evaluation metric for each body movement model, select a highest ranked body movement model based on the at least one metric calculated for each body movement model, combine the highest ranked body movement model with at least one other body movement model of the plurality of body movement models, calculate at least one evaluation metric for each combination of body movement models, and determine a highest ranked combination of body movement models to predict a bodily expression of emotion.

Abstract: A touchscreen includes a touch drive electrode, a touch sensing electrode, a touch drive circuit, a gate drive circuit, a touch drive electrode lead, and a liquid crystal panel. The touch drive electrode extends along a first direction. The touch drive circuit is disposed in a non-display area of the touchscreen, and an extension direction of the touch drive circuit is parallel to an extension direction of the touch drive electrode. The extension direction of the touch drive circuit is the same as the extension direction of the touch drive electrode. Each touch drive electrode is connected to the touch drive circuit using a touch drive electrode lead. The touch drive electrode lead is disposed in an area in which a gate drive circuit is located. The touch drive electrode lead and the gate drive circuit are disposed in an overlapping manner and are isolated from each other.

Abstract: A display mura correction method includes: obtaining initial image data of an initial image displayed on a display of a terminal, where the initial image data includes image data of a mura region corresponding to the initial image, when at least one pixel whose brightness value is not in a reference threshold set exists in the initial image data, the mura region is a region that is covered by the at least one pixel on the display, and a target feature data set of the mura region includes a set of target feature data of the at least one pixel; obtaining, according to a correspondence between feature data and a correction value and the target feature data of the at least one pixel, a target correction value set corresponding to the target feature data set; and writing the target correction value set into the terminal.

Abstract: The architectural complexity of a neural network is reduced by selectively pruning channels. A cost metric for a convolution layer is determined. The cost metric indicates a resource cost per channel for the channels of the layer. Training the neural network includes, for channels of the layer, updating a channel-scaling coefficient based on the cost metric. The channel-scaling coefficient linearly scales the output of the channel. A constant channel is identified based on the channel-scaling coefficients. The neural network is updated by pruning the constant channel. Model weights are updated via a stochastic gradient descent of a training loss function evaluated on training data. The channel-scaling coefficients are updated via an iterative-thresholding algorithm that penalizes a batch normalization loss function based on the cost metric for the layer and a norm of the channel-scaling coefficients.

Abstract: A method of improving quality of crowdsourced affective data based on agreement relationship between a plurality of annotators include receiving, by a processor, a collection of stimuli previously given affective labels by the plurality of annotators, executing, by a processor, an algorithm operative to perform the steps including constructing an agreement multigraph as a probabilistic model including a pair-wise status of agreement between the affective labels given by different ones of the plurality of annotators, learning the probabilistic model computationally using the crowdsourced affective data, identifying a reliability of each of the plurality of annotators based on the learned model, and adjusting the crowdsourced affective data by calculating the affective labels of each stimuli based on the identified reliability of each of the plurality of annotators, thereby improving the quality of the crowdsourced affective data.

Abstract: Computationally efficient accelerated D2-clustering algorithms are disclosed for clustering discrete distributions under the Wasserstein distance with improved scalability. Three first-order methods include subgradient descent method with re-parametrization, alternating direction method of multipliers (ADMM), and a modified version of Bregman ADMM. The effects of the hyper-parameters on robustness, convergence, and speed of optimization are thoroughly examined. A parallel algorithm for the modified Bregman ADMM method is tested in a multi-core environment with adequate scaling efficiency subject to hundreds of CPUs, demonstrating the effectiveness of AD2-clustering.

Abstract: A display mura correction method includes: obtaining initial image data of an initial image displayed on a display of a terminal, where the initial image data includes image data of a mura region corresponding to the initial image, when at least one pixel whose brightness value is not in a reference threshold set exists in the initial image data, the mura region is a region that is covered by the at least one pixel on the display, and a target feature data set of the mura region includes a set of target feature data of the at least one pixel; obtaining, according to a correspondence between feature data and a correction value and the target feature data of the at least one pixel, a target correction value set corresponding to the target feature data set; and writing the target correction value set into the terminal.

Abstract: Computationally efficient accelerated D2-clustering algorithms are disclosed for clustering discrete distributions under the Wasserstein distance with improved scalability. Three first-order methods include subgradient descent method with re-parametrization, alternating direction method of multipliers (ADMM), and a modified version of Bregman ADMM. The effects of the hyper-parameters on robustness, convergence, and speed of optimization are thoroughly examined. A parallel algorithm for the modified Bregman ADMM method is tested in a multi-core environment with adequate scaling efficiency subject to hundreds of CPUs, demonstrating the effectiveness of AD2-clustering.

Abstract: Disclosed in some examples are various modifications to the shape regression technique for use in real-time applications, and methods, systems, and machine readable mediums which utilize the resulting facial landmark tracking methods.

Abstract: Disclosed in some examples are various modifications to the shape regression technique for use in real-time applications, and methods, systems, and machine readable mediums which utilize the resulting facial landmark tracking methods.