Candidate derivation for affine merge mode in video coding

A method for video encoding, comprising:

• obtaining a first affine candidate and a second affine candidate associated with a current block;
• calculating a first set of affine model parameters associated with one or more control point motion vectors (CPMVs) of the first affine candidate;
• calculating a second set of affine model parameters associated with one or more CPMVs of the second affine candidate;
• and
• performing a similarity check between the first affine candidate and the second affine candidate based on the first set of affine model parameters and the second set of affine model parameters.

An apparatus for video decoding or encoding, comprising:

• one or more processors;
• and
• a memory coupled to the one or more processors and configured to store instructions executable by the one or more processors,
• wherein the one or more processors, upon execution of the instructions, are configured to perform operations comprising:
• obtaining a first affine candidate and a second affine candidate associated with a current block;
• calculating a first set of affine model parameters associated with one or more control point motion vectors (CPMVs) of the first affine candidate;
• calculating a second set of affine model parameters associated with one or more CPMVs of the second affine candidate;
• and
• performing a similarity check between the first affine candidate and the second affine candidate based on the first set of affine model parameters and the second set of affine model parameters.