The present disclosure provides a method and an apparatus for motion vector refinement. An exemplary method includes: determining a plurality of first blocks associated with a first motion vector and a plurality of second blocks associated with a second motion vector; determining a sum of absolute transformed difference (SATD) between one of the plurality of first blocks and one of the plurality of second blocks; and refining the first motion vector and the second motion vector based on the determined SATDs.

In the present disclosure, a method of decoding a video signal and a device therefor are disclosed. Specifically, a method of decoding an image based on an inter prediction mode includes deriving a motion vector of an available spatial neighboring block around a current block; deriving a collocated block of the current block based on the motion vector of the spatial neighboring block; deriving a motion vector in a sub-block unit in the current block based on a motion vector of the collocated block; and generating a prediction block of the current block using the motion vector derived in the sub-block unit, wherein the collocated block may be specified by the motion vector of the spatial neighboring block in one pre-defined reference picture.

The present principles relate to a method and device. A method for encoding a residual block comprises: obtaining (500) a first coding mode relative to a first 2D transform when coding the residual blocks according to a coding mode relative to a first 2D transform is enabled; obtaining (510) a second coding mode relative to a second 2D transform when coding the residual blocks according to a coding mode relative to a second 2D transform is enabled; and encoding (530) the residual block according to either said first coding mode or said second coding mode or both; the method is characterized in that enabling or disabling (520) the coding of the residual block according to said second coding mode depends on said first coding mode. The present principles relate also to a method and device for encoding/decoding a picture.

Disclosed herein is video encoding or decoding for efficiently encoding video. The techniques of the present disclosure are related to various split shapes of a block, syntaxes representing various split types of blocks, and syntax elements represented at a high level therefor.

Block-based, low latency rate control for an encoding system in which a wavelet transform decomposes pixel blocks into subbands stored as subbands in wavelet blocks (WBs) for encoding. Quantization parameters (QPs) for the subbands in each WB are estimated using a method that minimizes wavelet-inverse distortion given a rate bound. For each subband, a rate curve is generated based on an unquantized DCT histogram and bit count statistics for the subband, and a distortion curve is generated based on the unquantized DCT histogram and a distortion estimate for the subband that is estimated using a masked estimator. Once the rate-distortion curves for the subbands are generated, a bisection search may be used to find a point on each curve where the slope is the same for all the curves. The QPs associated with those equally sloped points are the global minimizing QPs for the wavelet block.

Method, device, apparatus, and computer-readable storage medium to determine whether video block is a fractional boundary video block (See paragraph [0032] and FIG. 7.) and to partition the fractional boundary video block into inferred partitions using a subset of available partition modes (See paragraph [0033] and FIG. 8.) are disclosed.

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.

Methods of encoding and decoding for video data are describe in which significance maps are encoded and decoded using non-spatially-uniform partitioning of the map into parts, wherein the bit positions within each part are associated with a given context. Example partition sets and processes for selecting from amongst predetermined partition sets and communicating the selection to the decoder are described.

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 reconstructed picture on the basis of the residual sample for the current block, wherein the residual information may be derived via different syntax elements depending on whether a transform has been applied to the current block.

Disclosed according to one exemplary embodiment includes not limited to: a filtering unit configured to generate filtering information by filtering a residual image corresponding to a difference between an original image and a prediction image; an inverse filtering unit configured to generate inverse filtering information by inversely filtering the filtering information; an estimator configured to generate the prediction image based on the original image and reconstruction information; a CNN-based in-loop filter configured to receive the inverse filtering information and the prediction image and to output the reconstruction information; and an encoder configured to perform encoding based on the filtering information and information of the prediction image, and wherein the CNN-based in-loop filter is trained for each of the plurality of artefact sections according to an artefact value or for each of the plurality of quantization parameter sections according to a quantization parameter.