Abstract: A tiled display device includes two panels and two cover layers respectively disposed on the two panels. The two cover layers include a contact region. A top portion and a bottom portion of the contact region have a height H. One of the two cover layers has a thickness Tn. One of the two panels has a distance Xn between an upper surface of the one of the two panels and the bottom portion of the contact region. The one of the two panels is corresponding to the one of the two cover layers. The height H, the thickness Tn and the distance Xn satisfy the equation: 0<H/(Xn+Tn)<0.8.

Abstract: A filament forming method includes: performing first stage to apply first bias including gate and drain voltages to a resistive memory unit plural times until read current reaches first saturating state, latching read current in first saturating state as saturating read current, determining whether rate of increase of saturating read current is less than first threshold value; when rate of increase of saturating read current is not less than first threshold value, performing second stage to apply second bias, by increasing gate voltage and decreasing drain voltage, to the resistive memory unit plural times until read current reaches second saturating state, latching read current in second saturating state as saturating read current and determining whether rate of increase of saturating read current is less than first threshold value; finishing the method when rate of increase of saturating read current is less than first threshold value and saturating read current reaches target current value.

Abstract: A startup stage protection device used in Electric Multiple Unit (EMU) train coupler experiment is provided between two test cars. The startup stage protection device is arranged between and tightly abuts two test cars at the starting stage of the experiment to receive a compressing force in place of the coupler. The startup stage protection device separates from the two test cars after the end of the starting stage.

Abstract: One embodiment provides an apparatus comprising a memory stack including multiple memory dies and a parallel processor including a plurality of multiprocessors. Each multiprocessor has a single instruction, multiple thread (SIMT) architecture, the parallel processor coupled to the memory stack via one or more memory interfaces. At least one multiprocessor comprises a multiply-accumulate circuit to perform multiply-accumulate operations on matrix data in a stage of a neural network implementation to produce a result matrix comprising a plurality of matrix data elements at a first precision, precision tracking logic to evaluate metrics associated with the matrix data elements and indicate if an optimization is to be performed for representing data at a second stage of the neural network implementation, and a numerical transform unit to dynamically perform a numerical transform operation on the matrix data elements based on the indication to produce transformed matrix data elements at a second precision.

Abstract: An apparatus to facilitate optimization of a convolutional neural network (CNN) is disclosed. The apparatus includes optimization logic to receive a CNN model having a list of instructions and including pruning logic to optimize the list of instructions by eliminating branches in the list of instructions that comprise a weight value of 0.

Abstract: A forming operation of resistive memory device is provided. The operation includes: applying a pre-forming gate voltage and a pre-forming bit line voltage to a target memory cell; performing a dense switching forming operation, wherein the dense switching forming operation includes alternately performing dense set operations and dense reset operations on the target memory cell, wherein the dense set operation includes applying a dense switching gate voltage and a dense set bit line voltage; and performing a normal set operation on the target memory cell, wherein the normal set operation includes applying a normal set gate voltage and a normal set bit line voltage to the target memory cell, the normal set gate voltage is greater than the pre-forming gate voltage and the dense switching gate voltage, and the normal set bit line voltage is less than the pre-forming bit line voltage and the dense set bit line voltage.

Abstract: Methods and systems for budgeted and simplified training of deep neural networks (DNNs) are disclosed. In one example, a trainer is to train a DNN using a plurality of training sub-images derived from a down-sampled training image. A tester is to test the trained DNN using a plurality of testing sub-images derived from a down-sampled testing image. In another example, in a recurrent deep Q-network (RDQN) having a local attention mechanism located between a convolutional neural network (CNN) and a long-short time memory (LSTM), a plurality of feature maps are generated by the CNN from an input image. Hard-attention is applied by the local attention mechanism to the generated plurality of feature maps by selecting a subset of the generated feature maps. Soft attention is applied by the local attention mechanism to the selected subset of generated feature maps by providing weights to the selected subset of generated feature maps in obtaining weighted feature maps.

Abstract: A driving assistance system and a driving assistance method are provided. The driving assistance system includes a plurality of image capturing devices, a congestion calculation module, and a determination module. The image capturing devices capture a plurality of regional images of a plurality of road sections of the target lane. The congestion calculation module calculates a plurality of congestion levels corresponding to the road sections according to the regional images. The determination module provides a lane changing message to the vehicle according to the congestion levels.

Abstract: Provided herein are a preparation method and use of an anti-MET-and-RON bispecific antibody and an antibody-drug conjugate thereof. The anti-MET-and-RON bispecific antibody includes an anti-MET antibody fragment and an anti-RON antibody fragment which are linked to each other through a specific chemical “knobs-into-holes” structure.

Abstract: A recipe verifying method, a recipe verifying server, and a smart manufacturing controlling system using the same are provided. The recipe verifying method includes the following steps. An inputting recipe having a plurality of inputting parameters is intercepted. The inputting recipe is transmitted from one of the inspection apparatuses. A target best known method (BKM) recipe is searched out from a plurality of candidate BKM recipes according to the inputting recipe. A plurality of predetermined limitations are obtained according to the target BKM recipe. Whether the inputting parameters of the inputting recipe meet the predetermined limitations is determined. An error report is generated, if the inputting parameters of the inputting recipe do not meet the predetermined limitations. At least one error is heighted in the error report.

Abstract: The present invention includes unique, isolated monoclonal antibodies that bind the PSI domain of human RON (RONPSI), and methods for making and using the same and novel combination therapies of anti-RONPSI antibodies, including humanized anti-RONPSI antibodies, nucleic acids that encode the heavy chain, the light chain, or both the light and heave chains of the antibody, host cells, expression vectors, and wherein the antibody or binding fragments thereof can be further conjugated with one or more cytotoxic or chemotherapeutic agents.

Abstract: Provided herein are a preparation method and use of an anti-MET-and-RON bispecific antibody and an antibody-drug conjugate thereof. The anti-MET-and-RON bispecific antibody includes an anti-MET antibody fragment and an anti-RON antibody fragment which are linked to each other through a specific chemical “knobs-into-holes” structure.

Abstract: An online antifouling ship ballast water treatment system includes a ballast water main pipeline, a ballast water branch pipeline, a dispensing pipeline, a filter on the ballast water main pipeline, an electrolysis unit on the ballast water branch pipeline, and a dispensing pump on the dispensing pipeline. Some of seawater introduced into the system is filtered by the filter and flows to the ballast water branch pipeline and the electrolysis unit, some of seawater electrolyzed by the electrolysis unit is reinjected to the filter through the dispensing pump and the dispensing pipeline after a ballast process is completed, and kept in the filter until a next ballast operation, so that growth and propagation of marine organisms in the filter can be inhibited by sodium hypochlorite contained in a TRO solution when a ballast pump stops working. The present invention also provides a ship ballast water treatment method.

Abstract: Techniques for implementing zone redundant computing services in distributed computing systems are described herein. In one embodiment, a computing device is configured to query a database in a first availability zone for a list of computing services corresponding to a user account and provided by servers in a second availability zone different than the first availability zone. The computing device can then selects one or more computing services provided by the servers in the second availability zone from the queried list of computing services and transmit an instruction to the servers in the second availability zone for performing the requested data operation.

Abstract: The present invention provides a dehydrogenation tank. An atomization spray is disposed at the center of the upper part of the dehydrogenation tank, and flow stirring modules used for stirring a flowing solution are respectively disposed at the middle and the bottom of the dehydrogenation tank. Each of the two flow stirring modules includes at least two layers of flow stirring meshes. By means of disposing the atomizing spray and the flow stirring modules in the dehydrogenation tank, the TRO solution flowing into the dehydrogenation tank are fully stirred, so that hydrogen gas mixed with the TRO solution is able to diffuse out fully and rapidly, thereby increasing dehydrogenation efficiency as well as reducing volume of the dehydrogenation tank. In addition, the present invention also provides a ballast water treatment system having the dehydrogenation tank.

Abstract: A staple advance device includes a receptacle having a guide pathway formed by a low level surface, a high level surface, a tilted surface located between the low and the high level surfaces, and an inclined surface formed and located beside the low level surface and communicating with the low level surface, a barrel is slidably engaged in the receptacle, a guide plate is attached to the barrel for supporting the fasteners, an actuating lever is pivotally attached to the barrel with a pivot axle and includes a guide pin slidably engaged in the guide pathway of the receptacle, and a guide rod for feeding the fastener into the barrel selectively, a guide device includes a gate for engaging with the tilted surface and for forming an endless guide pathway for the receptacle selectively.

Abstract: Techniques for implementing zone redundant computing services in distributed computing systems are described herein. In one embodiment, a computing device is configured to query a database in a first availability zone for a list of computing services corresponding to a user account and provided by servers in a second availability zone different than the first availability zone. The computing device can then selects one or more computing services provided by the servers in the second availability zone from the queried list of computing services and transmit an instruction to the servers in the second availability zone for performing the requested data operation.

Abstract: The present invention includes unique, isolated monoclonal antibodies that bind the PSI domain of human RON (RONPSI), and methods for making and using the same and novel combination therapies of anti-RONPSI antibodies, including humanized anti-RONPSI antibodies, nucleic acids that encode the heavy chain, the light chain, or both the light and heave chains of the antibody, host cells, expression vectors, and wherein the antibody or binding fragments thereof can be further conjugated with one or more cytotoxic or chemotherapeutic agents.

Abstract: Aspects of the disclosure provide a method for merging compressed access units according to compression rates and/or positions of the respective compressed access units. The method can include receiving a sequence of compressed access units corresponding to a sequence of raw access units partitioned from an image or a video frame and corresponding to a sequence of memory spaces in a frame buffer, determining a merged access unit including at least two consecutive compressed access units based on compression rates and/or positions of the sequence of compressed access units. The merged access unit is to be stored in the frame buffer with a reduced gap between the at least two consecutive compressed access units compared with storing the at least two consecutive compressed access units in corresponding memory spaces in the sequence of memory spaces.

Abstract: VLSI distributed LC resonant clock networks having reduced inductor dimensions as well as simplified decoupling capacitances that are obtained by including one or more compensation capacitors. A compensation capacitor can be added in parallel with a clock capacitance and/or in parallel with a clock inductor. The presence of a compensation capacitance reduces the overhead associated with the inductor and the decoupling capacitor. The compensation capacitor (s) can be selectively switched into the network to create scalable resonant frequencies.