Abstract: Disclosed herein is an encoding method. The encoding method includes extracting a feature map from an input image, determining an encoding feature map based on the extracted feature map, generating a converted feature map by performing conversion on the encoding feature map, and performing encoding on the converted feature map.

Abstract: An artificial neural network system of the present invention can include an input layer that receives input data; an output layer that outputs output data corresponding to the input data; and a hidden layer group including a multiple number of hidden layers that are positioned between the input layer and the output layer and are connected by connection weights, with the connection weights updated based on a difference between the output data and prediction data corresponding to the input data, where each of the hidden layers can include a hidden node that includes an activation function, which can output 0 if the input value is negative and can determine the output value based on linear functions having different slopes according to the range of the input value if the input value is positive, and the linear functions can have smaller slopes for greater distances from the origin point.

Abstract: A super-resolution method according to the inventive concept may include receiving an input image of low resolution, separating the input image into low resolution (LR) unit patches, classifying a texture type of each of pixels included in the input image using a local binary pattern for the LR unit patches, generating high resolution (HR) unit patches corresponding to each of the pixels based on a mapping kernel corresponding to the texture type, and combining the HR unit patches based on a predetermined setting to generate an output image of high resolution.

Abstract: A method of rendering a stereoscopic 3D image from a single image, including receiving a collection of pairs of 3D images including an input image and an output image that is a 3D pair of the input image, training a neural network composed of convolutional layers without any fully connected layers, with pairs of 3D images from the collection, to receive an input image and generate an output image that is a 3D pair of the input image, wherein the neural network is provided as an application on a computing device, receiving an input image and generating an output image that is a 3D pair of the input image by the neural network.

Abstract: A super-resolution method according to the inventive concept may include receiving an input image of low resolution, separating the input image into low resolution (LR) unit patches, classifying a texture type of each of pixels included in the input image using a local binary pattern for the LR unit patches, generating high resolution (HR) unit patches corresponding to each of the pixels based on a mapping kernel corresponding to the texture type, and combining the HR unit patches based on a predetermined setting to generate an output image of high resolution.

Abstract: An artificial neural network system of the present invention can include an input layer that receives input data; an output layer that outputs output data corresponding to the input data; and a hidden layer group including a multiple number of hidden layers that are positioned between the input layer and the output layer and are connected by connection weights, with the connection weights updated based on a difference between the output data and prediction data corresponding to the input data, where each of the hidden layers can include a hidden node that includes an activation function, which can output 0 if the input value is negative and can determine the output value based on linear functions having different slopes according to the range of the input value if the input value is positive, and the linear functions can have smaller slopes for greater distances from the origin point.

Abstract: A method of rendering a stereoscopic 3D image from a single image, including receiving a collection of pairs of 3D images including an input image and an output image that is a 3D pair of the input image, training a neural network composed of convolutional layers without any fully connected layers, with pairs of 3D images from the collection, to receive an input image and generate an output image that is a 3D pair of the input image, wherein the neural network is provided as an application on a computing device, receiving an input image and generating an output image that is a 3D pair of the input image by the neural network.

Abstract: Disclosed are a data migration apparatus and method. The data migration apparatus includes a location information acquirer configured to acquire location information corresponding to one or more pieces of migration target data from a first database; a backup data restorer configured to receive backup data of data stored in the first database from a backup device, which stores the backup data, and restore the received backup data; and a data migrator configured to extract the migration target data from the restored backup data by using the acquired location information and migrate the extracted migration target data to a second database.

Abstract: A method for controlling speakers in a mobile station comprises detecting a movement of the mobile station. The method also comprises differentially controlling sound output by the speakers based on the detected movement of the mobile station. The differentially controlling the sound output by the speakers based on the detected movement of the mobile station may comprise at least one of adjusting a sound volume and selecting a sound effect. The detecting the movement of the mobile station may comprise detecting a tilt of the mobile station.

Abstract: A method for controlling speakers in a mobile station comprises detecting a movement of the mobile station. The method also comprises differentially controlling sound output by the speakers based on the detected movement of the mobile station. The differentially controlling the sound output by the speakers based on the detected movement of the mobile station may comprise at least one of adjusting a sound volume and selecting a sound effect. The detecting the movement of the mobile station may comprise detecting a tilt of the mobile station.

Abstract: A semiconductor memory device includes a first data scramble circuit, which is configured between a data input buffer and a memory cell block, for outputting data by inverting or maintaining a polarity of an input data in response to a data scramble control signal and a second data scramble circuit, which is configured between the memory cell block and a data output buffer, for outputting data by inverting or maintaining a polarity of an output data in response to a data scramble control signal.

Abstract: Disclosed is a semiconductor memory device capable of improving a timing margin in a logic circuit using a signal latch. The semiconductor memory device according to the present invention includes an address generator including a latch circuit, wherein the address generator receives an external address signal from an external circuit and latches the external address signal in response to a control signal; and a Y-predecoder for receiving outputs from the address generator and for generating a pulse in response to the control signal. Accordingly, the semiconductor memory device according to the present invention transfers a signal at safety without considering a separate timing margin, by using the control signal in an adjacent circuit as an enable signal.