Abstract: A method and an apparatus are disclosed for video coding using a nonrectangular block splitting structure prediction and transform accordingly. The video coding method and the apparatus additionally apply a nonrectangular block splitting structure, in addition to an existing splitting structure, to a current block. The video coding method and the apparatus generate reference samples according to the applied splitting structure and effectively perform prediction and transform on split blocks.

Abstract: Embodiments of the present specification provide a method and a device for processing a video signal for intra prediction. A method for decoding a video signal according an embodiment of the present disclosure includes: obtaining intra prediction information of a current block to which the intra prediction is applied, in which the intra prediction information includes a first flag indicating a predetermined first mode; determining an intra prediction mode of the current block based on the intra prediction information, wherein when the first flag corresponds to a first value, the first mode is determined as the intra prediction mode and when the first flag corresponds to a second value, one of a plurality of candidate modes is determined as the intra prediction mode; and generating a prediction sample of the current block based on the intra prediction mode and a reference sample in a current picture.

Abstract: Provided is a method of generating a language model using crossmodal information. The method includes: receiving language-based first modality information and non-language-based second modality information; converting the first modality information into a first byte sequence; converting the second modality information into a second byte sequence; converting the first and second byte sequences into a first embedding vector and a second embedding vector by applying an embedding technique for each modality; generating semantic association information between first and second modality information by inputting the first and second embedding vectors to a crossmodal transformer; and learning the language model by setting the generated semantic association information as training data.

Abstract: A method and a device for encoding/decoding a video are provided. A method for decoding a video according to the present disclosure includes deriving a most probable mode (MPM) list based on intra prediction modes of neighboring blocks adjacent to a current block. The method for decoding the video also includes generating prediction pixels by applying candidate modes in the MPM list or the intra prediction modes of the neighboring blocks to reference pixels of a first region adjacent to the current block. The method for decoding the video also includes calculating a sum of absolute transformed difference (SATD) between the prediction pixels and reconstructed pixels of the first region. The method for decoding the video also includes deriving a first intra prediction mode, based on the SATD. The method for decoding the video also includes deriving an intra prediction mode of the current block, based on the first intra prediction mode.

Abstract: A method and an apparatus are disclosed for video coding using an improved in-loop filter. The video coding method and the apparatus generate a residual frame from a reconstructed frame using a deep learning model. The video coding method and the apparatus improve performance of an in-loop filter by approximating an original residual frame by applying the generated residual frame to a linear model.

Abstract: An image decoding method performed by a decoding apparatus according to the present disclosure comprises the steps of: obtaining intra prediction information including at least one of MPM flag information, MRL index information, and planner flag information from a bitstream; checking the MPM flag information; checking the MRL index information on the basis of a case where the MPM flag information indicates that the MRL index information is checked; checking the planner flag information on the basis of a case where a value of the MRL index information is 0; deriving an intra prediction mode for a current block on the basis of the check for the planner flag information; deriving prediction samples for the current block on the basis of the intra prediction mode for the current block; and generating a reconstructed picture on the basis of the prediction samples.

Abstract: Some embodiments provide a method for synchronizing state between multiple smart NICs of a host computer that perform operations using dynamic state information. At a first smart NIC of the plurality of smart NICs, the method stores a set of dynamic state information. The method synchronizes the set of dynamic state information across a communication channel that connects the smart NICs so that each of the smart NICs also stores the set of dynamic state information.

Abstract: An image encoding/decoding method and apparatus are provided. An image decoding method performed by an image decoding apparatus includes determining a splitting structure of a current block composed of chroma components, determining a splitting type of the current block based on the splitting structure, obtaining a plurality of lower-layer blocks by splitting the current block based on the splitting type, and decoding the lower-layer blocks based on a prediction mode of the lower-layer blocks.

Abstract: An image decoding method according to this disclosure includes receiving image information including intra matrix-based intra prediction (MIP) flag information related to whether an MIP is applied to a current block through a bitstream, deriving an intra prediction type for the current block as the MIP based on the intra MIP flag information, generating intra prediction samples for the current block based on the MIP, and generating reconstructed samples for the current block based on the intra prediction samples.

Abstract: Provided is a pressure vessel intended for storing up a fluid having high pressure, and more specifically, to a pressure vessel capable of enabling a sealing pad made of an elastic material to effectively prevent the fluid from flowing out from a bonding portion of a nozzle and a liner. The pressure vessel includes: a liner formed to be hollow so as to enable its inside to be filled with the fluid; a nozzle combined with one side of the liner, or each end portion of both sides so as to enable a coupling part having a cylindrical shape to be located in the inside of the liner; a sealing pad having a ring-like shape configured to enable the coupling part to pass through its center, and formed of an elastic material; and a sealing nut coupled to the coupling part from a lower portion of the sealing pad.

Abstract: A sensor may include a reflective electrode, a photoelectric conversion layer on the reflective electrode and including one or more photoelectric conversion materials, a semi-transmissive electrode on the photoelectric conversion layer, a light transmitting buffer layer on the semi-transmissive electrode, and a semi-transmissive auxiliary layer on the light transmitting buffer layer.

Abstract: The present disclosure relates to a cleaning agent composition for a substrate for a semiconductor device and a method for cleaning a substrate for a semiconductor device using the same. The cleaning agent composition contains a silicon-based compound represented by Formula 1 and an aprotic organic solvent with a dielectric constant of 10 or less, which can form a surface protective film capable of preventing collapse of the pattern even in a wet cleaning process of fine patterns with high aspect ratios, thereby providing a method for manufacturing a semiconductor device with an improved semiconductor manufacturing yield.

Abstract: A method and an apparatus are disclosed for video coding using improved cross-component linear model (CCLM) prediction. The video coding method and the apparatus generate a first predictor of a current chroma block according to CCLM prediction, and additionally generate a second predictor of the current chroma block based on the neighboring pixels of the current chroma block in intra prediction of the current chroma block. The video coding method and the apparatus perform weighted combination on the first predictor and the second predictor.

Abstract: A method and an apparatus are disclosed for video coding for refining predicted signals of intra prediction based on deep learning. The video coding method and the apparatus generate a refined intra predicted signal to improve video coding efficiency and video quality. The video coding method and the apparatus adaptively input a prediction block according to intra prediction of a current block, block information related to the current block, and neighboring reconstructed reference samples into a deep learning network.

Abstract: An image encoding/decoding method and apparatus are provided. An image decoding method performed by an image decoding apparatus includes determining a splitting type of a current block, splitting the current block into a plurality of lower-layer blocks based on the splitting type, and decoding the lower-layer blocks. In this case, the current block may be a chroma block, and the determining the splitting type of the current block may be performed by disallowing a predetermined type, in which a width or height of the lower-layer blocks is a predetermined value, among a plurality of splitting types.

Abstract: An image decoding method according to the present document comprises the steps of: receiving a bitstream including residual information; deriving a quantized transform coefficient for a current block on the basis of the residual information included in the bitstream; deriving a residual sample for the current block on the basis of the quantized transform coefficient; and generating a restoration picture on the basis of the residual sample for the current block, wherein the residual information includes position information for indicating the position of a transform coefficient, which is not the last 0 in a transform block, and the position information can be derived if a transform skip is not applied to the transform block.

Abstract: Provided is an optical system which may acquire a hyperspectral image by acquiring a spectral image of an object to be measured, which includes, to collect spectral data and train the neural network, an image forming part forming an image from an object to be measured and transmitting collimated light, a slit moving to scan the incident image and passing and outputting a part of the formed image, and a first optical part obtaining spectral data by splitting light of the image received through the slit by wavelength. Also, the system includes, to decompose overlapped spectral data and to infer hyperspectral image data through the trained neural network, an image forming part forming an image from an object to be measured and transmitting collimated light, and a first optical part obtaining spectral data by splitting light of the received image by wavelength.

Abstract: The method for decoding an image by a decoding device according to the present document may comprise the steps of: acquiring image information comprising partitioning information for a current block; determining, on the basis of the size of the current block, whether to partition the current block; partitioning the current block into subblocks on the basis of the partitioning information, if it is determined that the current block is to be partitioned, and decoding the subblocks; and not partitioning the current block, if it is determined that the current block is not to be partitioned, and decoding the current block, wherein, if the current block is a chroma block, and the size of the current block is at most the minimum size of a chroma block, then it is determined that the current block is not to be partitioned.

Abstract: A video decoding method performed by a decoding apparatus according to the present document comprises the steps of: receiving a bit stream including residual information of a current block; deriving a specific number of the number of context encoding bins for context syntax elements for a current sub-block of the current block; decoding the context syntax elements for the current sub-block included in the residual information on the basis of the specific number; deriving transform coefficients for the current sub-block on the basis of the decoded context syntax elements; deriving residual samples for the current block on the basis of the transform coefficients; and generating a reconstructed picture on the basis of the residual samples.

Abstract: An image decoding method performed by a decoding apparatus according to the present document comprises the steps of: obtaining a sign data hiding availability flag indicating whether or not sign data hiding is available for a current slice; obtaining a TSRC availability flag indicating whether or not TSRC is available for a transform skip block of the current slice; obtaining residual coding information about the transform skip block on the basis of the TSRC availability flag; deriving a residual sample for the transform skip block on the basis of the residual coding information; and generating a reconstruction picture on the basis of the residual sample, wherein the TSRC availability flag is obtained on the basis of the sign data hiding availability flag.