Methods and apparatus on prediction refinement with optical flow

A computing device, comprising:

• one or more processors;
• a non-transitory computer-readable storage medium storing instructions executable by the one or more processors, wherein the one or more processors are configured to:
• divide a video block of a current picture into multiple video subblocks, wherein at least one of the multiple video subblocks is associated with two motion vectors;
• determine a first reference picture and a second reference picture associated with the two motion vectors of the at least one of the multiple video subblocks, wherein the first reference picture is before the current picture and the second reference picture is after the current picture in display order;
• determine first prediction samples of the video subblock from the first reference picture;
• determine second prediction samples of the video subblock from the second reference picture;
• determine horizontal and vertical gradient values of the first prediction samples and second prediction samples for bi-directional optical flow (BDOF) when the BDOF is applied;
• determine horizontal and vertical gradient values of the first prediction samples and second prediction samples for prediction refinement with optical flow (PROF) when the PROF is applied;
• determine motion refinements for samples in the video subblock according to the horizontal and vertical gradient values for the BDOF when the BDOF is applied;
• determine motion refinements for samples in the video subblock according to the horizontal and vertical gradient values for the PROF when the PROF is applied;
• and
• determine prediction samples of the video block based on the motion refinements;
• wherein the one or more processors are further configured to:
• derive extended prediction samples along left and right boundaries of first and second prediction blocks of the video subblock by copying from integer reference samples that are closest to respective fractional prediction samples in a horizontal direction;
• and
• derive extended prediction samples along top and bottom boundaries of first and second prediction blocks of the video subblock by copying from integer reference samples that are closest to respective fractional prediction samples in a vertical direction.

A method for transmitting a bitstream, comprising:

• generating the bitstream by performing a method for video encoding;
• and
• transmitting the bitstream,
• wherein the method for video encoding comprises:
• dividing a video block of a current picture into multiple video subblocks, wherein at least one of the multiple video subblocks is associated with two motion vectors;
• determining a first reference picture and a second reference picture associated with the two motion vectors of the at least one of the multiple video subblocks, wherein the first reference picture is before the current picture and the second reference picture is after the current picture in display order;
• determining first prediction samples of the video subblock from the first reference picture;
• determining second prediction samples of the video subblock from the second reference picture;
• determining horizontal and vertical gradient values of the first prediction samples and second prediction samples for bi-directional optical flow (BDOF) when the BDOF is applied;
• determining horizontal and vertical gradient values of the first prediction samples and second prediction samples for prediction refinement with optical flow (PROF) when the PROF is applied;
• determining motion refinements for samples in the video subblock according to the horizontal and vertical gradient values for the BDOF when the BDOF is applied;
• determining motion refinements for samples in the video subblock according to the horizontal and vertical gradient values for the PROF when the PROF is applied;
• and
• determining prediction samples of the video block based on the motion refinements;
• wherein the method for video encoding further comprises:
• deriving extended prediction samples along left and right boundaries of first and second prediction blocks of the video subblock by copying from integer reference samples that are closest to respective fractional prediction samples in a horizontal direction;
• and
• deriving extended prediction samples along top and bottom boundaries of first and second prediction blocks of the video subblock by copying from integer reference samples that are closest to respective fractional prediction samples in a vertical direction.