Methods and apparatus for adaptive motion vector candidate ordering for video encoding and decoding

1. A video encoder, comprising:

a processor; and

a memory wherein the processor is configured to:

encode a block in a picture by selecting an order of motion vector predictor candidates for the block, based on common information available at both the encoder and a corresponding decoder, wherein said common information comprise a consistency of the motion vector predictor candidates, a fidelity of the motion vector predictor candidates, and a reference frame index of the motion vector predictor candidates and is not transmitted in an encoded bitstream, and wherein the processor is responsive to a selection frequency of the motion vector predictor candidates in a number of already previously encoded blocks, and wherein candidates with higher frequency have smaller index.

2. The apparatus of claim 1 , wherein a category classification is performed to determine one of a plurality of categories to which the block belongs, and the motion vector candidate selection frequency is determined from the number of already encoded blocks that belong to the same one of the plurality of categories as the block.

3. The apparatus of claim 2 , wherein a criterion for the category classification is a block prediction type.

4. The apparatus of claim 2 , wherein a criterion for the category classification is a block partition type.

5. The apparatus of claim 2 , wherein a criterion for the category classification is a block location.

6. The apparatus of claim 1 , wherein the consistency is a function of a difference between the motion vector predictor candidates that are available at both the video encoder and the corresponding decoder.

7. The apparatus of claim 1 , wherein the fidelity is a function of corresponding reconstructed residue coefficients which are available at both the video encoder and the corresponding decoder.

8. The apparatus of claim 1 , wherein the motion vector predictor candidates comprise a first motion vector predictor candidate for a first component of a motion vector and a second motion vector predictor candidate for a second component of the motion vector, and an order of the second motion vector predictor candidate for the second component is adapted responsive to a predictor index of the first component.

9. In a video encoder, a method, comprising:

encoding a block in a picture by selecting ( 315 , 520 , 715 , 910 , 1115 , 1325 ) an order of motion vector predictor candidates for the block, based on common information available at both the encoder and a corresponding decoder, wherein said common information comprise a consistency of the motion vector predictor candidates, a fidelity of the motion vector predictor candidates, and a reference frame index of the motion vector predictor candidates and is not transmitted in an encoded bitstream, and wherein the processor is responsive to a selection frequency of the motion vector predictor candidates in a number of already previously encoded blocks, and wherein candidates with higher frequency have smaller index.

10. The method of claim 9 , wherein a category classification is performed to determine one of a plurality of categories to which the block belongs, and the motion vector candidate selection frequency is determined from the number of already encoded blocks that belong to the same one of the plurality of categories as the block ( 520 ).

11. The method of claim 10 , wherein a criterion for the category classification is a block prediction type ( 515 ).

12. The method of claim 10 , wherein a criterion for the category classification is a block partition type ( 515 ).

13. The method of claim 10 , wherein a criterion for the category classification is a block location ( 515 ).

14. The method of claim 9 , wherein the consistency is a function of a difference between the motion vector predictor candidates that are available at both the video encoder and the corresponding decoder.

15. The method of claim 9 , wherein the fidelity is a function of corresponding reconstructed residue coefficients which are available at both the video encoder and the corresponding decoder.

16. The method of claim 9 , wherein the motion vector predictor candidates comprise a first motion vector predictor candidate for a first component of a motion vector and a second motion vector predictor candidate for a second component of the motion vector, and an order of the second motion vector predictor candidate for the second component is adapted responsive to a predictor index of the first component ( 1325 ).

17. A video decoder, comprising:

a processor; and

a memory wherein the processor is configured to:

decode a block in a picture by selecting an order of motion vector predictor candidates for the block, based on common information available at both the encoder and a corresponding decoder, wherein said common information comprise a consistency of the motion vector predictor candidates, a fidelity of the motion vector predictor candidates, and a reference frame index of the motion vector predictor candidates and is not transmitted in an encoded bitstream, and wherein the processor is responsive to a selection frequency of the motion vector predictor candidates in a number of already previously encoded blocks, and wherein candidates with higher frequency have smaller index.

18. The apparatus of claim 17 , wherein a category classification is performed to determine one of a plurality of categories to which the block belongs, and the motion vector candidate selection frequency is determined from the number of already encoded blocks that belong to the same one of the plurality of categories as the block.

19. The apparatus of claim 18 , wherein a criterion for the category classification is a block prediction type.

20. The apparatus of claim 18 , wherein a criterion for the category classification is a block partition type.

21. The apparatus of claim 18 , wherein a criterion for the category classification is a block location.

22. The apparatus of claim 17 , wherein the consistency is a function of a difference between the motion vector predictor candidates that are available at both the video decoder and the corresponding encoder.

23. The apparatus of claim 17 , wherein the fidelity is a function of corresponding reconstructed residue coefficients which are available at both the video decoder and the corresponding encoder.

24. The apparatus of claim 17 , wherein the motion vector predictor candidates comprise a first motion vector predictor candidate for a first component of a motion vector and a second motion vector predictor candidate for a second component of the motion vector, and an order of the second motion vector predictor candidate for the second component is adapted responsive to a predictor index of the first component.

25. In a video decoder, a method, comprising:

decoding a block in a picture by selecting ( 415 , 625 , 815 , 1015 , 1220 , 1425 ) an order of motion vector predictor candidates for the block, based on common information available at both the encoder and a corresponding decoder, wherein said common information comprise a consistency of the motion vector predictor candidates, a fidelity of the motion vector predictor candidates, and a reference frame index of the motion vector predictor candidates and is not transmitted in an encoded bitstream, and wherein the processor is responsive to a selection frequency of the motion vector predictor candidates in a number of already previously encoded blocks, and wherein candidates with higher frequency have smaller index.

26. The method of claim 25 , wherein a category classification is performed to determine one of a plurality of categories to which the block belongs, and the motion vector candidate selection frequency is determined from the number of already encoded blocks that belong to the same one of the plurality of categories as the block ( 625 ).

27. The method of claim 26 , wherein a criterion for the category classification is a block prediction type ( 620 ).

28. The method of claim 26 , wherein a criterion for the category classification is a block partition type ( 620 ).

29. The method of claim 26 , wherein a criterion for the category classification is a block location ( 620 ).

30. The method of claim 25 , wherein the consistency is a function of a difference between the motion vector predictor candidates that are available at both the video decoder and the corresponding encoder.

31. The method of claim 25 , wherein the fidelity is a function of corresponding reconstructed residue coefficients which are available at both the video decoder and the corresponding encoder.

32. The method of claim 25 , wherein the motion vector predictor candidates comprise a first motion vector predictor candidate for a first component of a motion vector and a second motion vector predictor candidate for a second component of the motion vector, and an order of the second motion vector predictor candidate for the second component is adapted responsive to a predictor index of the first component ( 1425 ).

33. A non-transitory computer readable tangible media having instructions stored thereon that when executed on a processor, cause the processor to perform steps, comprising:

a block in a picture encoded by selecting an order of motion vector predictor candidates for the block, based on common information available at both the encoder and a corresponding decoder, wherein said common information comprise a consistency of the motion vector predictor candidates, a fidelity of the motion vector predictor candidates, and a reference frame index of the motion vector predictor candidates and is not transmitted in an encoded bitstream, and wherein the processor is responsive to a selection frequency of the motion vector predictor candidates in a number of already previously encoded blocks, and wherein candidates with higher frequency have smaller index.