Apparatus and method to support encoding and decoding video data

An apparatus comprising:

• a memory having stored therein at least one received block of encoded video data;
• a control circuit operably coupled to the memory and configured to:
• determine a primary prediction mode to employ when decoding the block of encoded video data;
• infer at least one corresponding companion prediction mode from a companion prediction mode look-up table to correspond to the primary prediction mode;
• generate a mixed prediction for the block of encoded video data by mixing a prediction formed using the primary prediction mode with a prediction formed using the at least one companion prediction mode;
• and
• following the generating of the mixed prediction for the block of encoded video data, automatically update the companion prediction mode lookup table based on accuracy maps wherein each of the accuracy maps quantifies accuracy of a respective prediction mode at each position in the block and the automatic updating is in response to an instruction in a received bitstream that includes the block of encoded video data.

A method comprising:

• by a control circuit:
• determining a primary prediction mode to employ when at least one of encoding and decoding at least one block of video data;
• inferring from a companion prediction mode lookup table, a corresponding companion prediction mode to correspond to the primary prediction mode;
• generating a mixed prediction for the at least one block of video data by mixing a prediction formed using the primary prediction mode with a prediction formed using the companion prediction mode;
• following the generating of the mixed prediction for the at least one block of video data, automatically updating the companion prediction mode lookup table;
• wherein the companion prediction mode lookup table includes accuracy maps and wherein each of the accuracy maps identifies at what block positions a prediction mode performs well, and wherein the inferring of the corresponding companion prediction mode comprises, at least in part, comparing at least some of the accuracy maps for companion prediction modes with an accuracy map for the primary prediction mode.

A method comprising:

• determining a plurality of possible different primary prediction modes to employ when encoding the block of video data;
• for each primary prediction mode of the plurality of possible different primary prediction modes:
• determining, using a companion prediction mode lookup table, a corresponding companion prediction mode to correspond to that primary prediction mode,
• generating a respective mixed prediction for the block of video data by mixing a first respective prediction formed using that primary prediction mode with a second respective prediction formed using the corresponding companion prediction mode of that primary prediction mode, and
• subsequent to the generating of the respective mixed prediction for the block of video data, determining an efficacy metric for the respective mixed prediction, and automatically update the companion prediction mode lookup table;
• using the efficacy metrics to identify a selected primary prediction mode for a most efficacious one of the plurality of mixed predictions;
• and
• using the selected primary prediction mode to encode the block of video data.

A method comprising:

• determining a primary prediction mode to employ when decoding a block of encoded video data;
• inferring at least one corresponding companion prediction mode from a companion prediction mode look-up table to correspond to the primary prediction mode;
• generating a mixed prediction for the block of encoded video data by mixing a prediction formed using the primary prediction mode with a prediction formed using the at least one companion prediction mode;
• and
• subsequent to the generating of the mixed prediction for the block of encoded video data, automatically update the companion prediction mode lookup table based on accuracy maps wherein each of the accuracy maps quantifies accuracy of a respective prediction mode at each position in the block and the automatic updating is in response to an instruction in a received bitstream that includes the block of encoded video data.

An apparatus comprising:

• a memory having stored therein at least one received block of encoded video data;
• and
• a control circuit operably coupled to the memory and configured to:
• determine a primary prediction mode to employ when at least one of encoding and decoding the at least received one block of video data,
• infer from a companion prediction mode lookup table, a corresponding companion prediction mode to correspond to the primary prediction mode,
• generate a mixed prediction for the at least one block of video data by mixing a prediction formed using the primary prediction mode with a prediction formed using the companion prediction mode, and
• following the generating of the mixed prediction for the at least one block of video data, automatically updating the companion prediction mode lookup table,
• wherein the companion prediction mode lookup table includes accuracy maps and wherein each of the accuracy maps identifies block positions at which a prediction mode performs well, and wherein the inferring of the corresponding companion prediction mode comprises, at least in part, comparing at least some of the accuracy maps for companion prediction modes with an accuracy map for the primary prediction mode.