Abstract: A pouch type battery case for a secondary battery, which accommodates an electrode assembly, in which electrodes and separators are stacked, includes a first cup part and a second cup part, which are recessed in a pouch film, respectively; an accommodation part, which is provided in a longitudinal direction of edges of the first cup part and the second cup part, which face each other, between the first cup part and the second cup part and in which one side of the electrode assembly is configured to be accommodated; inclined parts extending from opposing ends, respectively, of the accommodation part and provided to be recessed in the pouch film; and line forming parts extending linearly from outer ends, respectively, of the respective inclined parts and provided to be recessed in the pouch film in the same direction as the longitudinal direction of the accommodation part.

Abstract: Provided are image decoding/encoding methods and apparatuses for increasing coding efficiency and reducing complexity while using a flexible tree split. The image decoding method includes: obtaining, from a bitstream, a bin string corresponding to a split shape mode of a coding unit; determining a split rule of the coding unit; and splitting the coding unit into a plurality of coding units, based on at least one of the split rule and the bin string corresponding to the split shape mode, wherein the determining of the split rule includes: determining an allowable first range of a size of the coding unit according to a ratio of a width and a height of the coding unit; and determining an allowable second range of the size of the coding unit according to the split shape mode of the coding unit.

Abstract: A refresh circuit is configured to generate a counting signal by counting a refresh command and to generate a plurality of preliminary refresh cycle change signals by decoding the counting signal. The refresh circuit is also configured to change a refresh cycle based on one of the plurality of preliminary refresh cycle change signals and to perform a refresh operation.

Abstract: A method and apparatus for performing transformation and inverse transformation on a current block by using multi-core transform kernels in video encoding and decoding processes. A video decoding method may include obtaining, from a bitstream, multi-core transformation information indicating whether multi-core transformation kernels are to be used according to a size of a current block; obtaining horizontal transform kernel information and vertical transform kernel information from the bitstream when the multi-core transformation kernels are used according to the multi-core transformation information; determining a horizontal transform kernel for the current block according to the horizontal transform kernel information; determining a vertical transform kernel for the current block according to the vertical transform kernel information; and performing inverse transformation on the current block by using the horizontal transform kernel and the vertical transform kernel.

Abstract: A method and apparatus for performing transformation and inverse transformation on a current block by using multi-core transform kernels in video encoding and decoding processes. A video decoding method may include obtaining, from a bitstream, multi-core transformation information indicating whether multi-core transformation kernels are to be used according to a size of a current block; obtaining horizontal transform kernel information and vertical transform kernel information from the bitstream when the multi-core transformation kernels are used according to the multi-core transformation information; determining a horizontal transform kernel for the current block according to the horizontal transform kernel information; determining a vertical transform kernel for the current block according to the vertical transform kernel information; and performing inverse transformation on the current block by using the horizontal transform kernel and the vertical transform kernel.

Abstract: A method and apparatus for performing transformation and inverse transformation on a current block by using multi-core transform kernels in video encoding and decoding processes. A video decoding method may include obtaining, from a bitstream, multi-core transformation information indicating whether multi-core transformation kernels are to be used according to a size of a current block; obtaining horizontal transform kernel information and vertical transform kernel information from the bitstream when the multi-core transformation kernels are used according to the multi-core transformation information; determining a horizontal transform kernel for the current block according to the horizontal transform kernel information; determining a vertical transform kernel for the current block according to the vertical transform kernel information; and performing inverse transformation on the current block by using the horizontal transform kernel and the vertical transform kernel.

Abstract: A method and apparatus for performing transformation and inverse transformation on a current block by using multi-core transform kernels in video encoding and decoding processes. A video decoding method may include obtaining, from a bitstream, multi-core transformation information indicating whether multi-core transformation kernels are to be used according to a size of a current block; obtaining horizontal transform kernel information and vertical transform kernel information from the bitstream when the multi-core transformation kernels are used according to the multi-core transformation information; determining a horizontal transform kernel for the current block according to the horizontal transform kernel information; determining a vertical transform kernel for the current block according to the vertical transform kernel information; and performing inverse transformation on the current block by using the horizontal transform kernel and the vertical transform kernel.

Abstract: Disclosed herein are image-based object recognition method and system by and in which a learning server performs image-based object recognition based on the learning of environment variable data. The image-based object recognition method includes: receiving an image acquired through at least one camera, and segmenting the image on a per-frame basis; primarily learning labeling results for one or more objects in the image segmented on a per-frame basis; performing primary reasoning by performing object detection in the image through a model obtained as a result of the primary learning; performing data labeling based on the results of the primary reasoning, and performing secondary learning with weights allocated to respective boxes obtained by the primary reasoning and estimated as object regions; and finally detecting one or more objects in the image through a model generated as a result of the secondary learning.

Abstract: Provided are an encoding method and apparatus, and a decoding method and apparatus for adaptively selecting a context model used to entropy-encode and entropy-decode a syntax element, based on various shapes of coding units. The image decoding method includes: determining a context model based on block shape information including at least one of a shape, direction, width, ratio of width and height, or size of a coding unit; obtaining, from a bitstream based on the context model, information about a split shape mode for splitting the coding unit; and determining a split shape mode of the coding unit, based on the information about the split shape mode.

Abstract: Disclosed herein are image-based object recognition method and system by and in which a learning server performs image-based object recognition based on the learning of environment variable data. The image-based object recognition method includes: receiving an image acquired through at least one camera, and segmenting the image on a per-frame basis; primarily learning labeling results for one or more objects in the image segmented on a per-frame basis; performing primary reasoning by performing object detection in the image through a model obtained as a result of the primary learning; performing data labeling based on the results of the primary reasoning, and performing secondary learning with weights allocated to respective boxes obtained by the primary reasoning and estimated as object regions; and finally detecting one or more objects in the image through a model generated as a result of the secondary learning.

Abstract: An apparatus for manufacturing a secondary battery includes: a recess induction part configured to allow a folding part folded in a battery case of a secondary battery to be recessed inward so as to provide a recess part when the battery case of the secondary battery is folded; a first guide, of which one side of one end is recessed to define a concave space, wherein, when the battery case of the secondary battery is folded, the one end is configured to be inserted into the battery case to push the folding part to an outside so that the recess induction part is configured to be inserted into the concave space; and a second guide configured to hold a sealing part of the battery case so as to fix the battery case.

Abstract: Provided are image decoding/encoding methods and apparatuses for increasing coding efficiency and reducing complexity while using a flexible tree split. The image decoding method includes: obtaining, from a bitstream, a bin string corresponding to a split shape mode of a coding unit; determining a split rule of the coding unit; and splitting the coding unit into a plurality of coding units, based on at least one of the split rule and the bin string corresponding to the split shape mode, wherein the determining of the split rule includes: determining an allowable first range of a size of the coding unit according to a ratio of a width and a height of the coding unit; and determining an allowable second range of the size of the coding unit according to the split shape mode of the coding unit.

Abstract: Provided are an encoding method and apparatus, and a decoding method and apparatus for adaptively selecting a context model used to entropy-encode and entropy-decode a syntax element, based on various shapes of coding units. The image decoding method includes: determining a context model based on block shape information including at least one of a shape, direction, width, ratio of width and height, or size of a coding unit; obtaining, from a bitstream based on the context model, information about a split shape mode for splitting the coding unit; and determining a split shape mode of the coding unit, based on the information about the split shape mode.

Abstract: Provided are image decoding/encoding methods and apparatuses for increasing coding efficiency and reducing complexity while using a flexible tree split. The image decoding method includes: obtaining, from a bitstream, a bin string corresponding to a split shape mode of a coding unit; determining a split rule of the coding unit; and splitting the coding unit into a plurality of coding units, based on at least one of the split rule and the bin string corresponding to the split shape mode, wherein the determining of the split rule includes: determining an allowable first range of a size of the coding unit according to a ratio of a width and a height of the coding unit; and determining an allowable second range of the size of the coding unit according to the split shape mode of the coding unit.

Abstract: Provided are an encoding method and apparatus, and a decoding method and apparatus for adaptively selecting a context model used to entropy-encode and entropy-decode a syntax element, based on various shapes of coding units. The image decoding method includes: determining a context model based on block shape information including at least one of a shape, direction, width, ratio of width and height, or size of a coding unit; obtaining, from a bitstream based on the context model, information about a split shape mode for splitting the coding unit; and determining a split shape mode of the coding unit, based on the information about the split shape mode.

Abstract: A semiconductor package includes a package substrate and a semiconductor chip mounted on the package substrate. The package substrate includes a signal bump land and an anchoring bump land, and the semiconductor chip includes a signal bump and an anchoring bump. The signal bump is bonded to the signal bump land, the anchoring bump is disposed to be adjacent to the anchoring bump land, and a bottom surface of the anchoring bump is located at a level which is lower than a top surface of the anchoring bump land with respect to a surface of the package substrate.

Abstract: A method and apparatus for performing transformation and inverse transformation on a current block by using multi-core transform kernels in video encoding and decoding processes. A video decoding method may include obtaining, from a bitstream, multi-core transformation information indicating whether multi-core transformation kernels are to be used according to a size of a current block; obtaining horizontal transform kernel information and vertical transform kernel information from the bitstream when the multi-core transformation kernels are used according to the multi-core transformation information; determining a horizontal transform kernel for the current block according to the horizontal transform kernel information; determining a vertical transform kernel for the current block according to the vertical transform kernel information; and performing inverse transformation on the current block by using the horizontal transform kernel and the vertical transform kernel.

Abstract: Provided is an artificial neural network learning apparatus for deep learning. The apparatus includes an input unit configured to acquire an input data or a training data, a memory configured to store the input data, the training data, and a deep learning artificial neural network model, and a processor configured to perform computation based on the artificial neural network model, in which the processor sets the initial weight depending on the number of nodes belonging to a first layer and the number of nodes belonging to a second layer of the artificial neural network model, and determines the initial weight by compensation by multiplying a standard deviation (?) by a square root of a reciprocal of a probability of a normal probability distribution for a remaining section except for a section in which an output value of the activation function converges to a specific value.

Abstract: An image decoding method includes: splitting a first block included in an image on the basis of at least one of a split type and a split direction of the first block to determine at least one second block from the first block; determining one of a prediction mode of the at least one second block and whether to split the at least one second block on the basis of at least one of a size and a shape of the determined at least one second block; obtaining a prediction block of a block included in the at least one second block on the basis of one of the determined prediction mode and whether to split the at least one second block; and restoring the block included in the at least one second block on the basis of the prediction block of the block included in the at least one second block. Here, the split type represents one of binary-split, tri-split, and quad-split.

Abstract: A Depth of Charge (DOC) setting apparatus includes a charging and discharging unit for completely charging a battery to a set target voltage and discharging the completely-charged battery; a profile obtaining unit for obtaining a voltage profile for capacity and voltage of the battery while the battery is charged and discharged and obtaining a differential profile for capacity and differential voltage of the battery from the obtained voltage profile; and a processor for sequentially selecting any one of a plurality of preset voltages and setting the selected preset voltage as the target voltage and for obtaining a feature value for a target peak in each of a plurality of differential profiles and setting a DOC for the battery based on the plurality of obtained feature values.