Systems and methods for model parameter optimization in three dimensional based color mapping
Systems, methods, and devices are disclosed for performing adaptive color space conversion and adaptive entropy encoding of LUT parameters. A video bitstream may be received and a first flag may be determined based on the video bitstream. The residual may be converted from a first color space to a second color space in response to the first flag. The residual may be coded in two parts separated by the most significant bits and least significant bits of the residual. The residual may be further coded based on its absolute value.
1. A method for encoding video content, the method comprising:
obtaining a current picture of a video;
identifying a temporal level of the current picture;
calculating a potential rate-distortion cost based on the identified temporal level of the current picture, wherein distortion is weighted more heavily for a lower temporal level than a higher temporal level;
determining a look-up table (LUT) size of a color mapping LUT for the current picture based on the potential rate-distortion cost;
generating the color mapping LUT of the determined LUT size for the current picture;
deriving, using the color mapping LUT of the LUT size, an inter-layer reference (ILR) picture based on the current picture; and
including an indication of the determined LUT size for the color mapping LUT associated with the current picture in a bitstream.
2. The method of claim 1 , wherein the LUT size for the color mapping LUT associated with the current picture is selected from a plurality of potential LUT sizes, and the method further comprises:
calculating a plurality of potential rate-distortion costs that correspond to coding the current picture with the plurality of potential LUT sizes; and
selecting an optimal LUT size that corresponds to the smallest rate-distortion cost as the LUT size for the color mapping LUT associated with the current picture.
3. The method of claim 2 , wherein the plurality of potential LUT sizes are limited by a maximum LUT size, and the method further comprises:
estimating a number of bits for coding the current picture; and
determining the maximum LUT size based on the estimated number of bits for coding the current picture.
4. The method of claim 1 , further comprising:
determining a distortion weighting factor based on the temporal level of the current picture, wherein the potential rate-distortion cost is calculated using the distortion weighting factor such that distortion is weighted more heavily when the temporal level of the current picture is a low temporal level than when the temporal level of the current picture is a high temporal level.
5. The method of claim 1 , further comprising:
generating an enhancement layer picture based on the current picture;
determining a usage of the ILR picture in predicting the enhancement layer picture; and
determining, based on the determined usage of the ILR picture, a distortion weighting factor for calculating the potential rate-distortion cost such that distortion is weighted more heavily when the ILR picture is used in predicting the enhancement layer picture.
6. The method of claim 1 , further comprising:
calculating a first rate-distortion cost associated of coding the current picture using a color mapping LUT of a first LUT size; and
calculating a second rate-distortion cost associated of coding the current picture using a color mapping LUT of a second LUT size, wherein the LUT size for the color mapping LUT associated with the current picture is determined based on comparing the first and second rate-distortion costs.
7. The method of claim 1 , further comprising:
determining whether to disable picture-level color mapping LUT update based on the temporal level of the current picture, wherein the picture-level color mapping LUT update is disabled when the temporal level of the current picture is high.
8. A video encoding device comprising:
a processor configured to:
obtain a current picture in a video;
identify a temporal level of the current picture;
calculate a potential rate-distortion cost based on the identified temporal level of the current picture, wherein distortion is weighted more heavily for a lower temporal level than a higher temporal level;
determine a look-up table (LUT) size of a color mapping LUT for the current picture based on a potential rate-distortion cost;
generate the color mapping LUT of the determined LUT size for the current picture;
derive, using the color mapping LUT of the LUT size, an inter-layer reference (ILR) picture based on the current picture; and
include an indication of the determined LUT size for the color mapping LUT associated with the current picture in a bitstream.
9. The video encoding device of claim 8 , wherein the LUT size for the color mapping LUT associated with the current picture is selected from a plurality of potential LUT sizes, the processor is further configured to:
calculate a plurality of potential rate-distortion costs that correspond to coding the current picture with the plurality of potential LUT sizes; and
select an optimal LUT size having the smallest rate-distortion cost as the LUT size for the color mapping LUT associated with the current picture.
10. The video encoding device of claim 9 , wherein the plurality of potential LUT sizes are limited by a maximum LUT size, the processor is further configured to:
estimate a number of bits for coding the current picture; and
determine the maximum LUT size based on the estimated number of bits for coding the current picture.
11. The video encoding device of claim 8 , wherein the processor is further configured to:
determine a distortion weighting factor based on the temporal level of the current picture, wherein the potential rate-distortion cost is calculated using the distortion weighting factor such that distortion is weighted more heavily when the temporal level of the current picture is a low temporal level than when the temporal level of the current picture is a high temporal level.
12. The video encoding device of claim 8 , wherein the processor is further configured to:
generate an enhancement layer picture based on the current picture;
determine a usage of the ILR picture in predicting the enhancement layer picture; and
determine, based on the determined usage of the ILR picture, a distortion weighting factor for calculating the potential rate-distortion cost such that distortion is weighted more heavily when the ILR picture is used in predicting the enhancement layer picture.
13. The video encoding device of claim 8 , wherein the processor is further configured to:
calculate a first rate-distortion cost associated of coding the current picture using a color mapping LUT of a first LUT size; and
calculate a second rate-distortion cost associated of coding the current picture using a color mapping LUT of a second LUT size, wherein the LUT size for the color mapping LUT associated with the current picture is determined based on comparing the first and second rate-distortion costs.
14. The video encoding device of claim 8 , wherein the processor is further configured to:
determine whether to disable picture-level color mapping LUT update based on the temporal level of the current picture, wherein the picture-level color mapping LUT update is disabled when the temporal level of the current picture is high.