# US 2018/0324454 A1 · Method and apparatus of inter prediction using average motion vector for video coding

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2018/0324454 A1
- Document ID: US20180324454A1
- Status: Published application
- Kind: A1
- Application number: 15/772,876
- Filed: 2016-10-28
- Published: 2018-11-08
- Earliest claimed: 2015-11-05
- Assignees: Mediatek Inc.
- Applicants: Mediatek Inc.
- Inventors: Jian-Liang LIN; Yi-Wen CHEN
- Web page: https://patentgrep.com/patent/US20180324454A1

## CPC

- H04N 19/52: by predictive encoding
- H04N 19/159: Prediction type, e.g. intra-frame, inter-frame or bidirectional frame prediction
- H04N 19/176: the region being a block, e.g. a macroblock
- H04N 19/91: Entropy coding, e.g. variable length coding [VLC] or arithmetic coding
- H04N 19/124: Quantisation
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION

## Abstract

A method and apparatus for video coding utilizing a motion vector predictor (MVP) for a motion vector (MV) for a block are disclosed. According to an embodiment, a mean candidate is derived from at least two candidates in the current candidate list. The mean candidate includes two MVs for the bi-prediction or one MV for the uni-prediction, and at least one MV of the mean candidate is derived as a mean of the MVs of said at least two candidates in one of list 0 and list 1. The mean candidate is added to the current candidate list to form a modified candidate list, and one selected candidate is determined as a MVP or MVPs from the modified candidate list, for current MV or MVs of the current block. The current block is then encoded or decoded in Inter, Merge, or Skip mode utilizing the MVP or MVPs selected.

## Claims

### Claim 1

A method of video coding utilizing a motion vector predictor (MVP) for a motion vector (MV) for a block coded in Inter, Merge, or Skip mode, wherein the MV is associated with one given block and one corresponding reference block in a given reference picture in a given reference list, the method comprising:
receiving input data associated with a current block in a current picture;
deriving a current candidate list for the current block based on MVs (motion vectors) associated with neighbouring blocks of the current block according to a pre-defined order, wherein each candidate in the current candidate list includes one MV (motion vector) associated with list 0 or list 1 for uni-prediction and each candidate includes two MVs (motion vectors) associated with both list 0 and list 1 for bi-prediction;
deriving a mean candidate from at least two candidates in the current candidate list, wherein the mean candidate includes two MVs for the bi-prediction or one MV for the uni-prediction, and at least one MV of the mean candidate is derived as a mean of the MVs of said at least two candidates in one of list 0 and list 1;
adding the mean candidate to the current candidate list to form a modified candidate list;
determining, from the modified candidate list, one selected candidate as a MVP (motion vector predictor) or MVPs (motion vector predictors) for a current MV or current MVs of the current block; and
encoding or decoding the current block in Inter, Merge, or Skip mode utilizing the MVP or MVPs, wherein a MVD (MV difference) or MVDs (MV differences) between the MVP or MVPs and the current MV or current MVs respectively along with prediction residuals of the current block are signalled for the current block coded in the Inter mode, wherein an index indicating the selected MVP in a Merge candidate list along with the prediction residuals of the current block are signalled for the current block coded in the Merge mode, and wherein the index indicating the selected MVP in the Merge candidate list is signalled for the current block coded in the Skip mode without the prediction residuals of the current block.

### Claim 2 (depends on 1)

The method of claim 1, wherein said at least two candidates correspond to two candidates designated as a first candidate and a second candidate.

### Claim 3 (depends on 2)

The method of claim 2, wherein both of the two MVs of the mean candidate are derived from the first candidate and the second candidate, wherein one of the two MVs of the mean candidate is derived as a first mean MV of the first candidate and the second candidate in list 0 and the other of the two MVs of the mean candidate is derived as a second mean MV of the first candidate and the second candidate in list 1.

### Claim 4 (depends on 2)

The method of claim 2, wherein both the first candidate and the second candidate have existing MV in a first list and the MV of the mean candidate in the first list is derived as a mean MV of the first candidate and the second candidate in the first list; and only the first candidate has existing MV in a second list and the MV of the mean candidate in the second list is set to the MV of the first candidate; and wherein the first list and the second list correspond to list 0 and list 1 respectively, or list 1 and list 0 respectively.

### Claim 5 (depends on 2)

The method of claim 2, wherein the mean of the MVs of the first candidate and the second candidate is calculated for only one of horizontal and vertical components of the mean candidate.

### Claim 6 (depends on 5)

The method of claim 5, wherein the other of horizontal and vertical components of the mean candidate is set to the other of horizontal and vertical components of either the first candidate or the second candidate.

### Claim 7 (depends on 1)

The method of claim 1, wherein said at least two candidates correspond to all spatial candidates, all temporal candidates, or all spatial candidates and all temporal candidates.

### Claim 8 (depends on 1)

The method of claim 1, wherein the mean of the MVs of said at least two candidates in one of list 0 and list 1 is calculated using “rounding half up”, “rounding half down”, “rounding toward zero”, “rounding away from zero” to fit in limited bit-depth representation.

### Claim 9 (depends on 1)

The method of claim 1, wherein mean candidates are derived for all candidate pairs pointing to a same reference picture in at least one list unless the current candidate list or the modified candidate list is full.

### Claim 10 (depends on 1)

The method of claim 1, wherein the mean candidate is inserted into the current candidate list in a pre-defined location.

### Claim 11 (depends on 1)

The method of claim 1, wherein the mean candidate is derived from a group of candidates pointing to a target reference picture.

### Claim 12 (depends on 1)

The method of claim 1, wherein the current block corresponds to a coding unit (CU).

### Claim 13 (depends on 1)

The method of claim 1, wherein size of the current candidate list or the modified candidate list for Merge mode is dependent on size of the current block.

### Claim 14

An apparatus of video coding utilizing a motion vector predictor (MVP) for a motion vector (MV) for a block coded in Inter, Merge, or Skip mode, wherein the MV is associated with one given block and one corresponding reference block in a given reference picture in a given reference list, the apparatus comprising one or more electronic circuits or processors configured to:
receive input data associated with a current block in a current picture;
derive a current candidate list for the current block based on MVs (motion vectors) associated with neighbouring blocks of the current block according to a pre-defined order, wherein each candidate in the current candidate list includes one MV (motion vector) associated with list 0 or list 1 for uni-prediction and each candidate includes two MVs (motion vectors) associated with both list 0 and list 1 for bi-prediction;
derive a mean candidate from at least two candidates in the current candidate list, wherein the mean candidate includes two MVs for the bi-prediction or one MV for the uni-prediction, and at least one MV of the mean candidate is derived as a mean of the MVs of said at least two candidates in one of list 0 and list 1;
add the mean candidate to the current candidate list to form a modified candidate list;
determine, from the modified candidate list, one selected candidate as a MVP (motion vector predictor) or MVPs (motion vector predictors) for a current MV or current MVs of the current block; and
encode or decode the current block in Inter, Merge, or Skip mode utilizing the MVP or MVPs, wherein a MVD (MV difference) or MVDs (MV differences) between the MVP or MVPs and the current MV or current MVs respectively along with prediction residuals of the current block are signalled for the current block coded in the Inter mode, wherein an index indicating the selected MVP in a Merge candidate list along with the prediction residuals of the current block are signalled for the current block coded in the Merge mode, and wherein the index indicating the selected MVP in the Merge candidate list is signalled for the current block coded in the Skip mode without the prediction residuals of the current block.

### Claim 15 (depends on 14)

The apparatus of claim 14, wherein said at least two candidates correspond to two candidates designated as a first candidate and a second candidate.

### Claim 16 (depends on 15)

The apparatus of claim 15, wherein both of the two MVs of the mean candidate are derived from the first candidate and the second candidate, wherein one of the two MVs of the mean candidate is derived as a first mean MV of the first candidate and the second candidate in list 0 and the other of the two MVs of the mean candidate is derived as a second mean MV of the first candidate and the second candidate in list 1.

### Claim 17 (depends on 15)

The apparatus of claim 15, wherein both the first candidate and the second candidate have existing MV in a first list and the MV of the mean candidate in the first list is derived as a mean MV of the first candidate and the second candidate in the first list; and only the first candidate has existing MV in a second list and the MV of the mean candidate in the second list is set to the MV of the first candidate; and wherein the first list and the second list correspond to list 0 and list 1 respectively, or list 1 and list 0 respectively.

### Claim 18 (depends on 15)

The apparatus of claim 15, wherein the mean of the MVs of the first candidate and the second candidate is calculated for only one of horizontal and vertical components of the mean candidate.

### Claim 19 (depends on 14)

The apparatus of claim 14, wherein said at least two candidates correspond to all spatial candidates, all temporal candidates, or all spatial candidates and all temporal candidates.

### Claim 20 (depends on 14)

The apparatus of claim 14, wherein the mean of the MVs of said at least two candidates in one of list 0 and list 1 is calculated using “rounding half up”, “rounding half down”, “rounding toward zero”, “rounding away from zero” to fit in limited bit-depth representation.
