Methods and devices for restriction on motion vector difference

An apparatus for encoding a current video block of a video bitstream into a coded video bitstream, the apparatus comprising:

• a memory storing instructions;
• and
• a processor in communication with the memory, wherein, when the processor executes the instructions, the processor is configured to cause the apparatus to:
• receive a video bitstream comprising a current video block,
• determine, based on the video bitstream, a motion vector difference (MVD) for the current video block, wherein the current video block is an inter coded block in a compound reference mode,
• determine whether the MVD is signaled for only one prediction direction,
• in response to determining that the MVD is signaled for only one prediction direction, restrict the MVD to a pre-defined direction, and
• encode, based on the MVD, the current video block into the bitstream,
• wherein:
• when the processor executes the instructions, the processor is configured to further cause the apparatus to determine, based on a mode for the current video block, whether a horizontal direction and a vertical direction has non-zero MVD,
• in response to the current video block being in a NEW_NEARMV mode, the MVD corresponds to the horizontal direction for a reference frame in a first reference frame list, and in response to the current video block being in a NEAR_NEWMV mode, the MVD corresponds to the vertical direction for a reference frame in a second reference frame list, or
• in response to the current video block being in a NEW_NEARMV mode, the MVD corresponds to the vertical direction for a reference frame in a first reference frame list, and in response to the current video block being in a NEAR_NEWMV mode, the MVD corresponds to the horizontal direction for a reference frame in a second reference frame list.

A method of processing video data, the method comprising:

• performing, by a device comprising a memory storing instructions and a processor in communication with the memory, a conversion between video data and a bitstream by:
• receiving the video data comprising a current video block;
• determining, based on the video data, a motion vector difference (MVD) for the current video block, wherein the current video block is an inter coded block in a compound reference mode;
• determining whether the MVD is signaled for only one prediction direction;
• in response to determining that the MVD is signaled for only one prediction direction, restrict the MVD to a pre-defined direction;
• and
• encoding, based on the MVD, the current video block into the bitstream,
• wherein:
• the method further comprises determining, based on a mode for the current video block, whether a horizontal direction and a vertical direction has non-zero MVD,
• in response to the current video block being in a NEW_NEARMV mode, the MVD corresponds to the horizontal direction for a reference frame in a first reference frame list, and in response to the current video block being in a NEAR_NEWMV mode, the MVD corresponds to the vertical direction for a reference frame in a second reference frame list, or
• in response to the current video block being in a NEW_NEARMV mode, the MVD corresponds to the vertical direction for a reference frame in a first reference frame list, and in response to the current video block being in a NEAR_NEWMV mode, the MVD corresponds to the horizontal direction for a reference frame in a second reference frame list.