High-level syntax control flags for template matching-related coding tools in video coding
A device for decoding video data comprises one or more processors configured to: obtain a syntax element from a bitstream that includes an encoded representation of the video data; determine, based on the syntax element, that a template-matching tool is enabled; based on the template-matching tool being enabled, applying the template-matching tool to generate a prediction block for a current coding unit (CU) of the video data; and reconstruct the current CU based on the prediction block for the current CU.
1 . A device for decoding video data, the device comprising:
a memory comprising one or more storage media, the memory configured to store the video data; and
one or more processors implemented in circuitry, the one or more processors configured to:
obtain a first syntax element from a bitstream that includes an encoded representation of the video data, wherein the first syntax element is a sequence-level syntax element, a picture-level syntax element, a slice-level syntax element, or a tile-level syntax element, wherein presence of syntax elements indicating whether intra template matching prediction (intraTMP), intra block copy template matching with advanced motion vector prediction (IBC-TM-AMVP), and intra block copy template matching with merge mode (IBC-TM-AMVP) are enabled is not dependent on the first syntax element;
determine, based on the first syntax element, that the bitstream includes one or more template-matching tool syntax elements indicating whether one or more template-matching tools are enabled, wherein the one or more template-matching tool syntax elements include a second syntax element;
obtain the second syntax element from the bitstream;
determine, based on the second syntax element, that a template-matching tool is enabled;
based on the template-matching tool being enabled, apply the template-matching tool to generate a prediction block for a current coding unit (CU) of the video data; and
reconstruct the current CU based on the prediction block for the current CU.
2 . The device of claim 1 , wherein the template-matching tool is one of:
template-matching advanced motion vector prediction (TM-AMVP),
geometric partitioning mode (GPM) split mode reordering,
candidate reordering for regular merge mode with motion vector difference (MMVD) and affine MMVD,
motion vector difference (MVD) sign prediction,
reference picture reordering,
template-matching merge mode,
template-matching GPM,
template-matching combined inter-intra prediction (CIIP),
adaptive re-ordering of merge candidates,
temporal motion vector prediction (TMVP) and non-adjacent merge candidate type reordering,
template-matching overlapped block motion compensation (TM-OBMC),
intra template matching prediction (IntraTMP),
intra-block copy-template matching-advanced motion vector prediction (IBC-TM-AMVP), or
intra-block copy-template matching-merge (IBC-TM-AMVP).
3 . The device of claim 1 , wherein:
the template-matching tool is a first template-matching tool, and
the one or more processors are further configured to determine, based on the second syntax element, whether the bitstream includes a third syntax element that indicates whether a second template-matching tool is enabled.
4 . The device of claim 3 , wherein the first template-matching tool is template-matching merge mode and the second template-matching tool is template-matching combined intra/inter prediction (CIIP) mode.
5 . The device of claim 3 , wherein:
the first template-matching tool is adaptive reordering of merge candidates (ARMC), and
the second template-matching tool is one of: geometric partitioning mode (GPM) split mode reordering, candidate reordering for regular merge mode with motion vector difference (MMVD) and affine MMVD, motion vector difference (MVD) sign prediction, reference picture reordering, temporal motion vector prediction (TMVP) or non-adjacent merge candidate type reordering.
6 . The device of claim 1 , wherein:
the one or more processors are further configured to:
based on the second syntax element indicating that intra-block copy template matching merge mode (IBC-TM-MRG) or intra-block copy template matching advance motion vector prediction (IBC-TM-AMVP) is disabled, set a pruning threshold to a value; and
prune a first candidate from a list based on a difference between a motion vector component of the first candidate and a motion vector component of a second candidate in the list being less than the pruning threshold, and
wherein to apply the template-matching tool to generate the prediction block for the current CU, the one or more processors are configured to apply the template-matching tool to generate the prediction block for the current CU based on a candidate in the list.
7 . The device of claim 1 , further comprising a display configured to display decoded video data.
8 . A method of decoding video data, the method comprising:
obtaining a first syntax element from a bitstream that includes an encoded representation of the video data, wherein the first syntax element is a sequence-level syntax element, a picture-level syntax element, a slice-level syntax element, or a tile-level syntax element, wherein presence of syntax elements indicating whether intra template matching prediction (intraTMP), intra block copy template matching with advanced motion vector prediction (IBC-TM-AMVP), and intra block copy template matching with merge mode (IBC-TM-AMVP) are enabled is not dependent on the first syntax element;
determining, based on the first syntax element, that the bitstream includes one or more template-matching tool syntax elements indicating whether one or more template-matching tools are enabled, wherein the one or more template-matching tool syntax elements include a second syntax element;
obtaining the second syntax element from the bitstream;
determining, based on the second syntax element, that a template-matching tool is enabled;
based on the template-matching tool being enabled, applying the template-matching tool to generate a prediction block for a current coding unit (CU) of the video data; and
reconstructing the current CU based on the prediction block for the current CU.
9 . The method of claim 8 , wherein the template-matching tool is one of:
template-matching advanced motion vector prediction (TM-AMVP),
geometric partitioning mode (GPM) split mode reordering,
candidate reordering for regular merge mode with motion vector difference (MMVD) and affine MMVD,
motion vector difference (MVD) sign prediction,
reference picture reordering,
template-matching merge mode,
template-matching GPM,
template-matching combined inter-intra prediction (CIIP),
adaptive re-ordering of merge candidates,
temporal motion vector prediction (TMVP) and non-adjacent merge candidate type reordering,
template-matching overlapped block motion compensation (TM-OBMC),
intra template matching prediction (IntraTMP),
intra-block copy-template matching-advanced motion vector prediction (IBC-TM-AMVP), or
intra-block copy-template matching-merge (IBC-TM-AMVP).
10 . The method of claim 8 , wherein:
the template-matching tool is a first template-matching tool, and
the method further comprises determining, based on the second syntax element, whether the bitstream includes a third syntax element that indicates whether a second template-matching tool is enabled.
11 . The method of claim 10 , wherein the first template-matching tool is template-matching merge mode and the second template-matching tool is template-matching combined intra/inter prediction (CIIP) mode.
12 . The method of claim 10 , wherein:
the first template-matching tool is adaptive reordering of merge candidates (ARMC), and
the second template-matching tool is one of: geometric partitioning mode (GPM) split mode reordering, candidate reordering for regular merge mode with motion vector difference (MMVD) and affine MMVD, motion vector difference (MVD) sign prediction, reference picture reordering, temporal motion vector prediction (TMVP) or non-adjacent merge candidate type reordering.
13 . The method of claim 8 , wherein:
the method further comprises:
based on the second syntax element indicating intra-block copy template matching merge mode (IBC-TM-MRG) or intra-block copy template matching advance motion vector prediction (IBC-TM-AMVP) is disabled, setting a pruning threshold to a value; and
pruning a first candidate from a list based on a difference between a motion vector component of the first candidate and a motion vector component of a second candidate in the list being less than the pruning threshold, and
applying the template-matching tool to generate the prediction block for the current CU comprises applying the template-matching tool to generate the prediction block for the current CU based on a candidate in the list.
14 . One or more non-transitory computer-readable storage media having instructions stored thereon that, when executed by one or more processors, cause the one or more processors to:
obtain a first syntax element from a bitstream that includes an encoded representation of video data, wherein the first syntax element is a sequence-level syntax element, a picture-level syntax element, a slice-level syntax element, or a tile-level syntax element, wherein presence of syntax elements indicating whether intra template matching prediction (intraTMP), intra block copy template matching with advanced motion vector prediction (IBC-TM-AMVP), and intra block copy template matching with merge mode (IBC-TM-AMVP) are enabled is not dependent on the first syntax element;
determine, based on the first syntax element, that the bitstream includes one or more template-matching tool syntax elements indicating whether one or more template-matching tools are enabled, wherein the one or more template-matching tool syntax elements include a second syntax element;
obtain the second syntax element from the bitstream;
determine, based on the second syntax element, that a template-matching tool is enabled;
based on the template-matching tool being enabled, apply the template-matching tool to generate a prediction block for a current coding unit (CU) of the video data; and
reconstruct the current CU based on the prediction block for the current CU.
15 . The one or more non-transitory computer-readable storage media of claim 14 , wherein the template-matching tool is one of:
template-matching advanced motion vector prediction (TM-AMVP),
geometric partitioning mode (GPM) split mode reordering,
candidate reordering for regular merge mode with motion vector difference (MMVD) and affine MMVD,
motion vector difference (MVD) sign prediction,
reference picture reordering,
template-matching merge mode,
template-matching GPM,
template-matching combined inter-intra prediction (CIIP),
adaptive re-ordering of merge candidates,
temporal motion vector prediction (TMVP) and non-adjacent merge candidate type reordering,
template-matching overlapped block motion compensation (TM-OBMC),
intra template matching prediction (IntraTMP),
intra-block copy-template matching-advanced motion vector prediction (IBC-TM-AMVP), or
intra-block copy-template matching-merge (IBC-TM-AMVP).
16 . The one or more non-transitory computer-readable storage media of claim 14 , wherein:
the template-matching tool is a first template-matching tool, and
execution of the instructions further causes the one or more processors to determine, based on the second syntax element, whether the bitstream includes a third syntax element that indicates whether a second template-matching tool is enabled.
17 . The one or more non-transitory computer-readable storage media of claim 16 , wherein the first template-matching tool is template-matching merge mode and the second template-matching tool is template-matching combined intra/inter prediction (CIIP) mode.
18 . The one or more non-transitory computer-readable storage media of claim 16 , wherein:
the first template-matching tool is adaptive reordering of merge candidates (ARMC), and
the second template-matching tool is one of: geometric partitioning mode (GPM) split mode reordering, candidate reordering for regular merge mode with motion vector difference (MMVD) and affine MMVD, motion vector difference (MVD) sign prediction, reference picture reordering, temporal motion vector prediction (TMVP) or non-adjacent merge candidate type reordering.
19 . The one or more non-transitory computer-readable storage media of claim 14 , wherein:
execution of the instructions further causes the one or more processors to:
based on the second syntax element indicating that intra-block copy template matching merge mode (IBC-TM-MRG) or intra-block copy template matching advance motion vector prediction (IBC-TM-AMVP) is disabled, set a pruning threshold to a value; and
prune a first candidate from a list based on a difference between a motion vector component of the first candidate and a motion vector component of a second candidate in the list being less than the pruning threshold, and
wherein to apply the template-matching tool to generate the prediction block for the current CU, the one or more processors are configured to apply the template-matching tool to generate the prediction block for the current CU based on a candidate in the list.