A macro block size determining unit 1 determines the size of each macro block on a frame-by-frame basis. A macro block dividing unit 2 divides an inputted image into macro blocks each having the size determined by the macro block size determining unit 1. A macro block coding unit 3 determines a coding mode for each of the macro blocks divided by the macro block dividing unit 2, and codes pixel values in each of the macro blocks in the determined coding mode.

An encoder/decoder is based on a combination of two audio or video channels to obtain a first combination signal as a mid-signal and a residual signal derivable using a predicted side signal derived from the mid-signal. A decoder uses the prediction residual signal, the first combination signal, a prediction direction indicator and prediction information to derive decoded first channel and second channel signals. A real-to-imaginary transform can be applied for estimating the imaginary part of the spectrum of the first combination signal. The prediction signal used in the derivation of the prediction residual signal, the real-valued first combination signal is multiplied by a real portion of the complex prediction information and the estimated imaginary part of the first combination signal is multiplied by an imaginary portion of the complex prediction information.

An encoder/decoder is based on a combination of two audio or video channels to obtain a first combination signal as a mid-signal and a residual signal derivable using a predicted side signal derived from the mid-signal. A decoder uses the prediction residual signal, the first combination signal, a prediction direction indicator and prediction information to derive decoded first channel and second channel signals. A real-to-imaginary transform can be applied for estimating the imaginary part of the spectrum of the first combination signal. The prediction signal used in the derivation of the prediction residual signal, the real-valued first combination signal is multiplied by a real portion of the complex prediction information and the estimated imaginary part of the first combination signal is multiplied by an imaginary portion of the complex prediction information.

According to the present invention, an inter-prediction method includes: receiving mode information on the inter-prediction of a current block; decoding the received mode information; and performing inter-prediction using the decoded mode information. According to the present invention, image compression efficiency may be improved.

According to the present invention, an inter-prediction method includes: receiving mode information on the inter-prediction of a current block; decoding the received mode information; and performing inter-prediction using the decoded mode information. According to the present invention, image compression efficiency may be improved.

An intra prediction method according to the present invention comprises the following steps: performing a directional prediction using at least one of a neighboring pixel of a current block and a left upper corner pixel positioned at a left upper corner of the current block so as to obtain a first prediction value for the current block; obtaining a second prediction value for the current block using the reference sample positioned in the current block; and weighted summing the first prediction value and the second prediction value using a weighting matrix so as to obtain a final prediction value for the current block. According to the present invention, image encoding/decoding efficiency may be improved.

An intra prediction method according to the present invention comprises the following steps: performing a directional prediction using at least one of a neighboring pixel of a current block and a left upper corner pixel positioned at a left upper corner of the current block so as to obtain a first prediction value for the current block; obtaining a second prediction value for the current block using the reference sample positioned in the current block; and weighted summing the first prediction value and the second prediction value using a weighting matrix so as to obtain a final prediction value for the current block. According to the present invention, image encoding/decoding efficiency may be improved.

Provided are an image encoding/decoding method and device. An image decoding method performed by an image decoding device according to the present disclosure includes the steps of: determining a quantization parameter of the current block on the basis of whether color space conversion is applied to a residual sample of the current block; determining a transform coefficient of the current block on the basis of the quantization parameter; determining the residual sample of the current block by using the transform coefficient; and resetting the value of the residual sample on the basis of whether the color space conversion is applied.

Provided are an image encoding/decoding method and device. An image decoding method performed by an image decoding device according to the present disclosure includes the steps of: determining a quantization parameter of the current block on the basis of whether color space conversion is applied to a residual sample of the current block; determining a transform coefficient of the current block on the basis of the quantization parameter; determining the residual sample of the current block by using the transform coefficient; and resetting the value of the residual sample on the basis of whether the color space conversion is applied.

A method and apparatus are provided for processing with a trained neural network, and for training of such neural network for image modification, which relate to image processing and in particular to modification of an image using the processing such as the neural network. The processing is performed to generate an output image. The output image is generated by processing the input image with the neural network. The processing with the neural network includes at least one stage including image down-sampling and filtering of the down-sampled image and at least one stage of image up-sampling. The image down-sampling is performed by applying a strided convolution. According to the application, efficiency of the neural network is increased, which may lead to faster learning and improved performance.