SbTMVP-based inter prediction method and apparatus

1. An image decoding method performed by a decoding apparatus, the method comprising:

deriving a reference block based on a motion shift;

determining availability of a subblock-based temporal motion vector predictor (sbTMVP) candidate for a current block, based on a prediction mode of the reference block;

deriving motion information for subblocks of the current block based on a case where the sbTMVP candidate is available;

generating predicted samples based on the motion information for the subblocks of the current block; and

generating a reconstructed picture based on the predicted samples, wherein the deriving the motion information for the subblocks of the current block comprises:

deriving a reference subblock for each of the subblocks in the current block based on the motion shift, in a reference picture including the reference block; and

deriving the motion information for the each of the subblocks in the current block based on whether the each of the reference subblocks is available,

wherein whether the each of the reference subblocks is available is determined based on a prediction mode of the each of the reference subblocks, and

wherein the each of the reference subblock is determined as unavailable based on the prediction mode of the each of the reference subblocks being a current picture referencing (CPR) mode.

2. The image decoding method of claim 1 , wherein based on a case where the prediction mode of the reference block is not an inter prediction mode or a first reference picture to which the reference block refers and a second reference picture in which the reference block is positioned are the same, it is determined that the sbTMVP candidate is not available.

3. The image decoding method of claim 1 , wherein based on a case where the prediction mode of the reference block is an inter prediction mode and a first reference picture to which the reference block refers and a second reference picture in which the reference block is positioned are not the same, it is determined that the sbTMVP candidate is available.

4. The image decoding method of claim 1 , wherein

based on a case where the prediction mode of a specific reference subblock among the reference subblocks is not an inter prediction mode, it is determined that the specific reference subblock is not available,

wherein motion information for a subblock in the current block corresponding to the unavailable specific reference subblock is derived as base motion information.

5. The image decoding method of claim 1 , further comprising: constructing a motion information candidate list for the current block,

wherein based on a case where the sbTMVP candidate is available, the sbTMVP candidate is included in the motion information candidate list, and

wherein based on a case where the sbTMVP candidate is not available, the sbTMVP candidate is excluded from the motion information candidate list.

6. The image decoding method of claim 1 , wherein the deriving of the reference block comprises:

deriving a motion vector from a spatial neighboring block of the current block;

performing a motion shift to a position indicated by the motion vector of the spatial neighboring block based on a position of a center sample of the current block; and

deriving the reference block based on the motion-shifted position.

7. The image decoding method of claim 6 , wherein the spatial neighboring block is a left neighboring block positioned at a left of the current block, and

wherein the motion vector of the spatial neighboring block is a motion vector derived from the left neighboring block based on whether the left neighboring block is available.

8. An image encoding method performed by an encoding apparatus, the method comprising:

deriving a reference block based on a motion shift;

determining availability of a subblock-based temporal motion vector predictor (sbTMVP) candidate for a current block, based on a prediction mode of the reference block;

deriving motion information for subblocks of the current block based on a case where the sbTMVP candidate is available;

generating predicted samples based on the motion information for the subblocks of the current block;

generating residual samples based on the predicted samples; and

encoding image information including information on the residual samples, wherein the deriving the motion information for the subblocks of the current block comprises:

deriving a reference subblock for each of the subblocks in the current block based on the motion shift, in a reference picture including the reference block; and

deriving the motion information for the each of the subblocks in the current block based on whether the each of the reference subblocks is available,

wherein whether the each of the reference subblocks is available is determined based on a prediction mode of the each of the reference subblocks, and

wherein the each of the reference subblock is determined as unavailable based on the prediction mode of the each of the reference subblocks being a current picture referencing (CPR) mode.

9. The image encoding method of claim 8 , wherein based on a case where the prediction mode of the reference block is not an inter prediction mode or a first reference picture to which the reference block refers and a second reference picture in which the reference block is positioned are the same, it is determined that the sbTMVP candidate is not available.

10. The image encoding method of claim 8 , wherein based on a case where the prediction mode of the reference block is an inter prediction mode and a first reference picture to which the reference block refers and a second reference picture in which the reference block is positioned are not the same, it is determined that the sbTMVP candidate is available.

11. The image encoding method of claim 8 ,

wherein based on a case where the prediction mode of a specific reference subblock among the reference subblocks is not an inter prediction mode, it is determined that the specific reference subblock is not available,

wherein motion information for a subblock in the current block corresponding to the unavailable specific reference subblock is derived as base motion information.

12. A non-transitory computer-readable storage medium storing encoded image information generated by an image encoding method, the image encoding method comprises:

deriving a reference block based on a motion shift;

determining availability of a subblock-based temporal motion vector predictor (sbTMVP) candidate for a current block, based on a prediction mode of the reference block;

deriving motion information for subblocks of the current block based on a case where the sbTMVP candidate is available;

generating predicted samples based on the motion information for the subblocks of the current block;

generating residual samples based on the predicted samples; and

encoding image information including information on the residual samples, wherein the deriving the motion information for the subblocks of the current block comprises:

deriving a reference subblock for each of the subblocks in the current block based on the motion shift, in a reference picture including the reference block; and

deriving the motion information for the each of the subblocks in the current block based on whether the each of the reference subblocks is available,

wherein whether the each of the reference subblocks is available is determined based on a prediction mode of the each of the reference subblocks, and

wherein the each of the reference subblock is determined as unavailable based on the prediction mode of the each of the reference subblocks being a current picture referencing (CPR) mode.