Method and apparatus for video coding

1. A method for video decoding in a decoder, comprising:

selecting, by processing circuitry of the decoder, a neighboring sub-block outside a current block for a current sub-block in the current block based on a prediction direction that indicates a spatial relationship between the current sub-block and the selected neighboring sub-block, the selected neighboring sub-block and the current block being in a coded picture from a coded bitstream, and a representative position of the current sub-block being on a reference line extended along the prediction direction from a reference position that is on or adjacent to a representative position of the selected neighboring sub-block;

determining, by the processing circuitry of the decoder, a reference motion vector corresponding to the reference position based on a motion vector for coding the selected neighboring sub-block;

determining, by the processing circuitry of the decoder, a motion vector predictor for the current sub-block based on the reference motion vector; and

decoding the current sub-block for output using inter-picture prediction performed based on the motion vector predictor.

2. The method of claim 1 , wherein the selected neighboring sub-block and the current sub-block are in a same column of the coded picture, and the selected neighboring sub-block is a top neighbor of the current block.

3. The method of claim 1 , wherein the selected neighboring sub-block and the current sub-block are in a same row of the coded picture, and the selected neighboring sub-block is a left neighbor of the current block.

4. The method of claim 1 , wherein a motion vector predictor for each remaining sub-block included in the current block is determined from a motion vector of a respective neighboring sub-block that is selected based on the prediction direction.

5. The method of claim 1 , further comprising:

decoding a codeword from the coded bitstream, the codeword indicating the prediction direction.

6. The method of claim 1 , further comprising:

decoding, from the coded bitstream, a residual motion vector for the current sub-block; and

combining the motion vector predictor with the residual motion vector to derive a current motion vector for the current sub-block.

7. The method of claim 1 , further comprising:

decoding at least one syntax element from the coded bitstream, the at least one syntax element indicating the prediction direction.

8. The method of claim 1 , further comprising:

deriving a motion vector for the current sub-block based on refining the motion vector predictor using one of template matching and bilateral matching.

9. The method of claim 1 , further comprising:

selecting another neighboring sub-block in the coded picture outside the current block for the current sub-block based on the prediction direction, the reference position being adjacent to another representative position of the other selected neighboring sub-block and the representative position of the selected neighboring sub-block; and

determining the reference motion vector corresponding to the reference position based on the motion vector for coding the selected neighboring sub-block and another motion vector for coding the other selected neighboring sub-block.

10. The method of claim 9 , wherein the determining the reference motion vector corresponding to the reference position is performed using a weighted combination of the motion vector for coding the selected neighboring sub-block and the other motion vector for coding the other selected neighboring sub-block.

11. A decoder device, comprising processing circuitry configured to:

select, a neighboring sub-block outside a current block for a current sub-block in the current block based on a prediction direction that indicates a spatial relationship between the current sub-block and the selected neighboring sub-block, the selected neighboring sub-block and the current block being in a coded picture from a coded bitstream, and a representative position of the current sub-block being on a reference line extended along the prediction direction from a reference position that is on or adjacent to a representative position of the selected neighboring sub-block;

determine a reference motion vector corresponding to the reference position based on a motion vector for coding the selected neighboring sub-block;

determine a motion vector predictor for the current sub-block based on the reference motion vector; and

decode the current sub-block for output using inter-picture prediction performed based on the motion vector predictor.

12. The decoder device of claim 11 , wherein the selected neighboring sub-block and the current sub-block are in a same column of the coded picture, and the selected neighboring sub-block is a top neighbor of the current block.

13. The decoder device of claim 11 , wherein the selected neighboring sub-block and the current sub-block are in a same row of the coded picture, and the selected neighboring sub-block is a left neighbor of the current block.

14. The decoder device of claim 11 , wherein the processing circuitry is further configured to:

determine a motion vector predictor for each remaining sub-block included in the current block from a motion vector of a respective neighboring sub-block that is selected based on the prediction direction.

15. The decoder device of claim 11 , wherein the processing circuitry configured to:

decode a codeword from the coded bitstream, the codeword indicating the prediction direction.

16. The decoder device of claim 11 , wherein the processing circuitry is further configured to:

select another neighboring sub-block in the coded picture outside the current block for the current sub-block based on the prediction direction, the reference position being adjacent to another representative position of the other selected neighboring sub-block and the representative position of the selected neighboring sub-block; and

determine the reference motion vector corresponding to the reference position based on the motion vector for coding the selected neighboring sub-block and another motion vector for coding the other selected neighboring sub-block.

17. The decoder device of claim 16 , wherein the processing circuitry is further configured to:

determine the reference motion vector corresponding to the reference position using a weighted combination of the motion vector for coding the selected neighboring sub-block and the other motion vector for coding the other selected neighboring sub-block.

18. The decoder device of claim 11 , wherein the processing circuitry is further configured to:

decode, from the coded bitstream, a residual motion vector for the current sub-block; and

derive a current motion vector for the current sub-block by combing the motion vector predictor for the current sub-block with the residual motion vector.

19. A non-transitory computer-readable storage medium storing a program executable by at least one processor to perform:

selecting a neighboring sub-block outside a current block for a current sub-block in the current block based on a prediction direction that indicates a spatial relationship between the current sub-block and the selected neighboring sub-block, the selected neighboring sub-block and the current block being in a coded picture from a coded bitstream, and a representative position of the current sub-block being on a reference line extended along the prediction direction from a reference position that is on or adjacent to a representative position of the selected neighboring sub-block;

determining a reference motion vector corresponding to the reference position based on a motion vector for coding the selected neighboring sub-block;

determining a motion vector predictor for the current sub-block based on the reference motion vector; and

decoding the current sub-block for output using inter-picture prediction performed based on the motion vector predictor.

20. The method of claim 1 , wherein the prediction direction is selected from a group of directions including a vertical prediction direction, a horizontal prediction direction, and a diagonal prediction direction.

21. The non-transitory computer-readable storage medium of claim 19 , wherein the prediction direction is selected from a group of directions including a vertical prediction direction, a horizontal prediction direction, and a diagonal prediction direction.

22. The decoder device of claim 11 , wherein the prediction direction is selected from a group of directions including a vertical prediction direction, a horizontal prediction direction, and a diagonal prediction direction.