Abstract: Aspects of the disclosure provide a method and an apparatus for video encoding. The apparatus includes processing circuitry configured to generate an initial feature representation from an input image to be encoded and perform an iterative update of values of a plurality of elements in the initial feature representation. The iterative update includes generate a coded representation corresponding to a final feature representation based on the final feature representation that has been updated from the initial feature representation by a number of iterations of the iterative update. A reconstructed image corresponding to the final feature representation is generated based on the coded representation. An encoded image corresponding to the final feature representation having updated values of the plurality of elements is generated. One of (i) a rate-distortion loss corresponding to the final feature representation or (ii) the number of iterations of the iterative update satisfies a pre-determined condition.

Abstract: A method of adaptive neural image compression with rate control by meta-learning includes receiving an input image and a hyperparameter; and encoding the received input image, based on the received hyperparameter, using an encoding neural network, to generate a compressed representation. The encoding includes performing a first shared encoding on the received input image, using a first shared encoding layer having first shared encoding parameters, performing a first adaptive encoding on the received input image, using a first adaptive encoding layer having first adaptive encoding parameters, combining the first shared encoded input image and the first adaptive encoded input image, to generate a first combined output, and performing a second shared encoding on the first combined output, using a second shared encoding layer having second shared encoding parameters.

Abstract: A near eye display system is provided. The near eye display system includes: a frame comprising a main body and two temple arms; at least one near eye sensor mounted on the main body and configured to measure user eye parameters; a first near eye display mounted on the main body and configured to form a first image projected on a first retina of a first eye; a second near eye display mounted on the main body and configured to form a second image projected on a second retina of a second eye; and a processing unit located at least at one of the two temple arms and configured to generate a display control signal based at least on the user eye parameters, wherein the display control signal drives the first near eye display and the second near eye display.

Abstract: A layout method includes disposing a first conductive path and a second conductive path across a boundary between a first layout device and a second layout device abutting the first layout device. The layout method also includes disposing a first cut layer on the first conductive path nearby the boundary, and disposing a second cut layer on the second conductive path nearby the boundary. The layout method also includes moving the first cut layer to align with the second cut layer.

Abstract: There is included a method and apparatus comprising computer code configured to cause a processor or processors to perform obtaining an input low resolution (LR) image comprising a height, a width, and a number of channels, implementing a feature learning deep neural network (DNN) configured to compute a feature tensor based on the input LR image, generating, by an upscaling DNN, a high resolution (HR) image, having a higher resolution than the input LR image, based on the feature tensor computed by the feature learning DNN, wherein a networking structure of the upscaling DNN differs depending on different scale factors, and wherein a networking structure of the feature learning DNN is a same structure for each of the different scale factors.

Abstract: Systems and methods for deep neural network (DNN)-based cross component prediction are provided. A method includes inputting a reconstructed luma block of an image or video into a DNN; and predicting, by the DNN, a reconstructed chroma block of the image or video based on the reconstructed luma block that is input. Luma and chroma reference information and side information may also be input into the DNN to predict the reconstructed chroma block. The various inputs may also be generated using processes such as downsampling and transformation.

Abstract: Method, apparatus, and non-transitory storage medium for end-to-end neural image compression using dependent scalar quantization with substitution, including receiving an input image; determining a substitute image based on the input image using a neural network based substitute feature generator; compressing the substitute image; quantizing the compressed substitute image to obtain a quantized representation of the input image with higher compression performance by using a first dependent scalar quantizer; and entropy encoding the substitute image using a neural network based encoder to generate a compressed representation of the quantized representation.

Abstract: Provided is an intelligent safe vehicle speed measurement method and an system capable of considering a state of a road surface, including: a laser scanning unit configured to obtain road surface textures and simulate contact under different vehicle loads and tire patterns; a wireless communication unit configured to obtain the model, load and tire information of incoming vehicles, offer a best-matching braking distance by search through database; a skid resistance prediction unit configured to obtain environmental parameters to correct the obtained best-matching braking distance, and offer a safe speed for incoming vehicles; and a warning reminder unit configured to broadcast the safe running speed to incoming vehicles. This system communicates between the vehicle and road, obtains the information from incoming vehicles in real time, and broadcasts the safe speed to incoming vehicles, ensuring the safety of the incoming vehicle during running.

Abstract: A memory device includes transistor structures and memory arc wall structures. The memory arc wall structures are embedded in the transistor structures. The transistor structure includes a dielectric column, a source electrode and a drain electrode, a gate electrode layer and a channel wall structure. The source electrode and the drain electrode are located on opposite sides of the dielectric column. The gate electrode layer is around the dielectric column, the source electrode, and the drain electrode. The channel wall structure is extended from the source electrode to the drain electrode and surrounds the dielectric column. The channel wall structure is disposed between the gate electrode layer and the source electrode, between the gate electrode layer, and the drain electrode, and between the gate electrode layer and the dielectric column. The memory arc wall structure is extended on and throughout the channel wall structure.

Abstract: A method of forming a ferroelectric random access memory (FeRAM) device includes: forming a layer stack over a substrate, where the layer stack includes alternating layers of a first dielectric material and a word line (WL) material; forming first trenches extending vertically through the layer stack; filling the first trenches, where filling the first trenches includes forming, in the first trenches, a ferroelectric material, a channel material over the ferroelectric material, and a second dielectric material over the channel material; after filling the first trenches, forming second trenches extending vertically through the layer stack, the second trenches being interleaved with the first trenches; and filling the second trenches, where filling the second trenches includes forming, in the second trenches, the ferroelectric material, the channel material over the ferroelectric material, and the second dielectric material over the channel material.

Abstract: A method of substitutional neural residual compression is performed by at least one processor and includes estimating motion vectors, based on a current image frame and a previous reconstructed image frame, obtaining a predicted image frame, based on the estimated motion vectors and the previous reconstructed image frame, and subtracting the obtained predicted image frame from the current image frame to obtain a substitutional residual. The method further includes encoding the obtained substitutional residual, using a first neural network, to obtain an encoded representation, and compressing the encoded representation.

Abstract: A multimedia information processing method, apparatus, electronic device, and medium are provided. The method includes: receiving a user's selection operation for any piece of multimedia information among multimedia information to be processed, the multimedia information to be processed comprising at least two pieces of multimedia information, any piece of multimedia information being any piece of multimedia information except the last piece; determining a target multimedia information piece on the basis of a selected operation; upon receiving a trigger operation for the target multimedia information piece, determining a corresponding processing method; on the basis of the determined processing method, correspondingly processing the target multimedia information piece. Thus, the complexity of processing multimedia information is reduced and efficiency is improved, thereby improving user experience.

Abstract: A hollow-core optical fiber may include a hollow core extending along a central longitudinal axis of the fiber; a substrate; a plurality of first cladding elements spaced apart from each other and positioned between the hollow core and the substrate, each of the first cladding elements extending in a direction parallel to the central longitudinal axis of the fiber, each of the first cladding elements including a first capillary; and a plurality of second cladding elements spaced apart from each other and positioned between the hollow core and the substrate, each of the second cladding elements extending in a direction parallel to the central longitudinal axis of the fiber, each of the second cladding elements including a second capillary. Each of the first cladding elements directly contacts the inner surface of the substrate, and none of the second cladding elements directly contacts the inner surface of the substrate.

Abstract: A hollow-core optical fiber may include a substrate having a tubular shape and an inner surface surrounding a central longitudinal axis of the hollow-core optical fiber; a hollow core extending through the substrate along the central longitudinal axis of the hollow-core optical fiber; and a plurality of cladding elements positioned between the central longitudinal axis of the hollow-core optical fiber and the substrate. Each of the plurality of cladding elements may extend in a direction parallel to the central longitudinal axis of the hollow-core optical fiber. Each of the plurality of cladding elements may include a wound glass sheet configured as a spiral, and each of the plurality of cladding elements may contact an interior surface of the substrate.

Abstract: A hollow-core optical fiber may include a hollow core extending along a central longitudinal axis of the fiber; a substrate, the substrate having a tubular shape and an inner surface surrounding the central longitudinal axis of the fiber; and a plurality of cladding elements positioned between the hollow core and the substrate, each of the cladding elements extending in a direction parallel to the central longitudinal axis of the fiber. Each of the cladding elements includes a primary capillary, the primary capillary directly contacting the inner surface of the substrate and having an inner surface defining a cavity, and a plurality of nested capillaries positioned within the cavity, each of the nested capillaries directly contacting the inner surface of the primary capillary.

Abstract: A transmission processing method includes a first communications device sending transmission assistance information of first data, or discarding the first data according to transmission configuration information. The transmission assistance information is used for assisting a second communications device to execute transmission processing of the first data.

Abstract: A method for producing a hollow-core preform may include rolling a glass sheet to form a rolled-glass structure; and attaching one or more of the rolled-glass structures to an inner surface of an annular support structure to form a hollow-core preform, wherein the inner surface of the annular support structure defines an interior cavity and the one or more of the rolled-glass structures are positioned within the interior cavity. The hollow-core preform may be drawn into a hollow-core optical fiber.

Abstract: Aspects of the disclosure provide a method, an apparatus, and a non-transitory computer-readable storage medium for video decoding. The apparatus can include processing circuitry. The processing circuitry is configured to decode first neural network update information in a coded bitstream for a first neural network in the video decoder. The first neural network is configured with first pretrained parameters. The first neural network update information corresponds to a first block in an image to be reconstructed and indicates a first replacement parameter corresponding to a first pretrained parameter in the first pretrained parameters. The processing circuitry is configured to update the first neural network in the video decoder based on the first replacement parameter. The processing circuitry can decode the first block based on the updated first neural network for the first block.

Abstract: Systems, apparatus and methods are provided for temperature assisted non-volatile storage device management in a non-volatile storage system. In one embodiment, a non-volatile storage system may comprise a temperature sensor, a non-volatile storage device and a processor. The processor may be configured to obtain a read-out from the temperature sensor, generate a predicted real-time on-die temperature for the non-volatile storage device based on the read-out, generate an estimated threshold voltage for reading data stored in the non-volatile storage device based on the predicted real-time on-die temperature and conduct a local sweep of a reference voltage using the estimated threshold voltage as a starting point to obtain a final read reference voltage with a minimum read bit error rate.

Abstract: Provided are an Electro Static Discharge (ESD) circuit and a memory. The ESD circuit includes: a detection circuit and multiple electrostatic discharge circuits. The detection circuit includes at least one sub-detection circuit connected between a first power end and a second power end. Each sub-detection circuit generates a sub-trigger signal based on a voltage change between the first power end and the second power end. The multiple electrostatic discharge circuits are connected between the first power end and the second power end. The multiple electrostatic discharge circuits are configured to be turned on according to the one or more sub-trigger signals.