Abstract: Embodiments of 3D memory devices and methods for forming the same are disclosed. In an example, a 3D memory device includes a substrate, a peripheral circuit on the substrate, a memory stack including interleaved conductive layers and dielectric layers above the peripheral circuit, a first semiconductor layer above the memory stack, a second semiconductor layer above and in contact with the first semiconductor layer, a plurality of channel structures each extending vertically through the memory stack and the first semiconductor layer, and an insulating structure extending vertically through the memory stack, the first semiconductor layer, and the second semiconductor layer.

Abstract: A computer implemented method includes obtaining a first deep neural network (DNN) model trained on labeled real image data for a downstream vision task, obtaining a second DNN model trained on synthetic images created with random image parameter values for the downstream vision task, obtaining a third DNN model trained on the labeled real image data and the synthetic images for the downstream vision task, performing a forward pass execution of each model to generate a loss, backpropagating the loss to modify parameter values, and iterating the forward pass execution and backpropagating with images generated by the modified parameters to jointly train the models and optimize the parameters.

Abstract: A semiconductor device is provided. The semiconductor device includes a first wafer having an array transistor formed therein, and a second wafer having a capacitor structure formed therein. The semiconductor device also includes a bonding interface formed between the first wafer and second wafer that includes a plurality of bonding structures. The bonding structures are configured to couple the array transistor to the capacitor structure to form a memory cell.

Abstract: Aspects of the disclosure provide a semiconductor device. The semiconductor device includes a first stack of layers including a source connection layer and a second stack of layers including gate layers and insulating layers. The gate layers and the insulating layers are stacked alternatively upon the first stack of layers. Further, the semiconductor device includes channel structures that are formed along the first direction in the first stack of layers and the second stack of layers, and a gate line cut structure having a trench that cuts through the first stack of layers and the second stack of layers. The trench is filled with at least an insulating layer. The semiconductor device includes a support structure having a first portion that is disposed at a side of the gate line cut structure and extended from the side of the gate line cut structure and underneath the second stack of layers.

Abstract: A method for manufacturing a memory device is provided. The method includes depositing a floating gate electrode film over a semiconductor substrate; patterning the floating gate electrode film into at least one floating gate electrode having at least one opening therein; depositing a control gate electrode film over the semiconductor substrate to overfill the at least one opening of the floating gate electrode; and patterning the control gate electrode film into at least one control gate electrode over the floating gate electrode.

Abstract: The present invention discloses a food swallowing simulating device, and belongs to the field of structural design of bio-simulation machinery. The device comprises an upper jaw matrix, a tongue matrix, a transmission device and a driving device; the upper jaw matrix and the tongue matrix are arranged oppositely, and opposite surfaces are curved surfaces; a channel is reserved between the curved surfaces to form a bionic oral cavity; and the tongue matrix is connected with the driving device by the transmission device, so that the motion of the tongue matrix is controlled by the driving device. Transition from static food swallowing mechanism simulation to dynamic food swallowing mechanism simulation is realized, and thus, a blank in the market is made up.

Abstract: A display apparatus and a control method for controlling a display apparatus are provided. The display apparatus includes: a display panel including multiple rows of scan lines and multiple columns of data lines, the scan lines and the data lines defining multiple sub-pixel regions, a first view pixel and a second view pixel being provided in each sub-pixel region; and a view control device connected with display panel, the view control device configured to: identify the number of users on a user side of display panel; and control a driving voltage of the scan lines according to the number of the users, so that the display panel controls first view pixel and second view pixel according to driving voltage so as to display a first view mode or a second view mode. A visual angle in the first view mode is greater than a visual angle in the second view mode.

Abstract: The disclosure discloses a methane value online real-time monitoring system, which includes a laser tunable system portion, a gas component detection system portion and a data processing system portion. The laser tunable system portion includes a computer, laser drivers, tunable lasers, a laser beam combiner, a collimator, and a beam expander, which are connected in sequence. The gas component detection system portion includes a gas supply system pipeline, a hose, a valve, and a gas chamber. The data processing system portion includes a detector, lock-in amplifiers, an oscilloscope, and a data acquisition card, which are connected in sequence. In a working process of the methane value online real-time monitoring system, gas mass namely the methane value change can be monitored in real time on line by collecting natural gas introduced into a combustion chamber through a gas supply system pipeline of a natural gas engine, and the knocking problem related to rapid change of the gas mass is solved.

Abstract: A complex flow tube for fine scaling coating of a PVC material for an automobile includes a base fixed to a mechanical arm, and a pipeline connected to the base for delivering a PVC sealant; the base is detachably butted with an interface of a PVC gluing pump mounted on the mechanical arm; the PVC gluing pump delivers the PVC sealant through the pipeline to a part to be coated or sealed of the automobile. The complex flow tube may be combined with the metal 3D printing technology, so that the manufactured complex flow tube has the advantages of being convenient to use, simple in structure, high in strength, not liable to break, etc.

Abstract: A backlight source, a liquid crystal module, and a display device are provided. The backlight source includes a back plate and a plurality of support structures. The back plate includes a side wall and a bottom plate, and the side wall and the bottom plate define a first groove. A plurality of the support structures are arranged along a circumferential direction of the side wall and spaced apart from each other. The support structure is detachably connected to the side wall, and extends toward a center of the accommodation groove to form a support plane.

Abstract: A fuel cell power controller tracks load current and fuel cell output voltage, and alerts on excessive fuel cell ramp rate, so another power source can supplement the fuel cell and/or the load can be reduced. A power engineering process makes efficient use of available fuel cell power by ramping up power flow rapidly when power is available, while respecting the ramp rate and other power limitations of the fuel cell and safety limitations of the load. Power flow decreases after an alert indicating an electrical output limitation of the fuel cell. Permitted power flow increases in response to a power demand increase (actual or requested) from the load in the absence of the alert. Power flow may increase or decrease in a fixed amount, a proportional amount, or per a sequence. A power controller relay may trip open on a low fuel cell output voltage or high load current.

Abstract: A supporting structure includes: a first supporting module located below the flexible screen and used for supporting the flexible screen, which has a first opening; and a second supporting module, which is located at a position corresponding to the first opening, when an external force acts through the flexible screen, the second supporting module forms a deformation protrusion towards a direction away from the flexible screen; and when the external force is removed, the deformation of the second supporting module is removed. Supporting the flexible screen corresponding to the first opening by the second supporting module can reduce a probability that the flexible screen collapses at the position of the first opening, thereby making appearance of the flexible screen flat.

Abstract: The discussion relates to context-aware environments. One example can include inwardly-facing cameras positioned around a periphery of an environment that defines a volume. The example can also include sensors positioned relative to the volume and configured to communicate with a user device in the volume. The example can also include an ambient perception component configured to track user locations in the volume and to detect user gestures relative to objects in the volume, and responsive to receiving a query from the user's device, to supplement the query with information derived from the objects.

Abstract: The present application provides a display panel and a display device. The display panel includes an array substrate and a color filter substrate arranged oppositely, and a light source on a side of the color filter substrate facing the array substrate. The color filter substrate includes a first region and a second region. An orthogonal projection of the array substrate on the color filter substrate is in the first region, and the light source is in the second region.

Abstract: Methods, systems, and apparatus include computer programs encoded on a computer-readable storage medium, including a system that controls content distribution using a feedback loop. Content is distributed over multiple different online channels using a same initial selection value for distribution over each different online channel. An observed user actions required for distribution of the content over the multiple different online channels is received through a feedback loop and for multiple different distributions of the content. Based on the observed user actions received through the feedback loop, a predicted user action rate is determined for the multiple different distributions across the multiple different online channels. The selection value is adjusted based on a difference between the predicted user action rate and a reference distribution amount specified by a provider of the content. The content is distributed over the multiple different online channels using the adjusted selection value.

Abstract: Methods, systems, and apparatus include computer programs encoded on a computer-readable storage medium, including a system that controls content distribution using a feedback loop. Content is distributed over multiple different online channels using a same initial selection value for distribution over each different online channel. An observed user actions required for distribution of the content over the multiple different online channels is received through a feedback loop and for multiple different distributions of the content. Based on the observed user actions received through the feedback loop, a predicted user action rate is determined for the multiple different distributions across the multiple different online channels. The selection value is adjusted based on a difference between the predicted user action rate and a reference distribution amount specified by a provider of the content. The content is distributed over the multiple different online channels using the adjusted selection value.

Abstract: The disclosure relates to a method, electronic device, and non-transitory computer-readable medium for collecting fingerprints. The method includes detecting a touch operation on a fingerprint region; when the touch operation on the fingerprint region is detected, detecting a touch signal within the fingerprint region as a first touch signal; comparing the first touch signal with a preset report threshold to obtain a first comparison result; determining whether a fingerprint image on the fingerprint region is acquired based on the first comparison result; when the fingerprint image is acquired, analyzing the fingerprint image to obtain an analysis result; and outputting a prompt based on the analysis result.

Abstract: An online system develops a model to predict the identity of unknown users accessing the online system. The online system interacts with users who are known by the online system (e.g., because they are logged in), termed known users, and users who are unknown by the online system. The model attempts to predict the identity of unknown users. To train the model, a set of training data with training weights is generated. The training data includes a set of access events from known users. The set can include access events from unknown users who accessed the system and subsequently became identified (referred to as hindsight events). To account for a distribution in training data, the training data is applied to a scoring model to identify training data that resembles known events. A scaling model then scales the scores to generate training weights. The weights may be higher for access events with characteristics that resembles hindsight events.

Abstract: A computer device is provided that includes a processor configured to simulate a virtual environment based on a set of virtual environment parameters, and perform ray tracing to render a view of the simulated virtual environment. The ray tracing includes generating a plurality of rays for one or more pixels of the rendered view of the simulated virtual environment. The processor is further configured to determine sub-pixel data for each of the plurality of rays based on intersections between the plurality of rays and the simulated virtual environment, and store the determined sub-pixel data for each of the plurality of rays in an image file.

Abstract: A computer implemented method includes obtaining a first deep neural network (DNN) model trained on labeled real image data for a downstream vision task, obtaining a second DNN model trained on synthetic images created with random image parameter values for the downstream vision task, obtaining a third DNN model trained on the labeled real image data and the synthetic images for the downstream vision task, performing a forward pass execution of each model to generate a loss, backpropagating the loss to modify parameter values, and iterating the forward pass execution and backpropagating with images generated by the modified parameters to jointly train the models and optimize the parameters.