# US 2020/0007892 A1 · Method, apparatus and medium for decoding or encoding

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

- Number: US 2020/0007892 A1
- Document ID: US20200007892A1
- Status: Published application
- Kind: A1
- Application number: 16/225,982
- Filed: 2018-12-19
- Published: 2020-01-02
- Earliest claimed: 2018-06-29
- Assignees: Tencent America LLC
- Applicants: Tencent America LLC
- Inventors: Xin ZHAO; Xiang LI; Shan LIU
- Web page: https://patentgrep.com/patent/US20200007892A1

## CPC

- H04N 19/61: in combination with predictive coding
- H04N 19/105: Selection of the reference unit for prediction within a chosen coding or prediction mode, e.g. adaptive choice of position and number of pixels used for prediction
- H04N 19/122: Selection of transform size, e.g. 8x8 or 2x4x8 DCT; Selection of sub-band transforms of varying structure or type
- H04N 19/147: according to rate distortion criteria (rate-distortion as a criterion for motion estimation H04N19/567)
- 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 19/597: specially adapted for multi-view video sequence encoding
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION

## Abstract

A method of utilizing an 8-bit primary transform core matrix to decode compressed video or image data or encode uncompressed video or image data. The method may include determining whether to use a first transform core matrix that is of a first size type or a second transform core matrix that is of a second size type that is smaller than the first size type. When a result of the determination is to use the first transform core matrix, encoding or decoding the target file using the 8-bit primary transform core matrix that has the size that is 64-point or larger. If not, the method may include extracting the second transform core matrix of the second size type from the first transform matrix and encoding or decoding the target file using the extracted second transform core matrix.

## Claims

### Claim 1

A method of decoding or encoding, the method comprising:
receiving information regarding a target file for encoding or decoding, the target file for encoding or decoding being one of: a compressed video or image data file or an uncompressed video or image data file;
determining, for the encoding or decoding of the target file, whether to use a first transform core matrix that is a Versatile Video Coding (VVC) standard transform and that has a first size type or use a second transform core matrix that is of a second size type that is smaller than the first size type, wherein the first size type is the 8-bit primary transform core matrix, which is represented by 8-bits and that has a size that is 64-point or larger;
when a result of the determination is to use the first transform core matrix of the first size type, constructing the first transform core matrix using integers used to construct the second transform core matrix and additional integers, and causing or transmitting information that causes the target file for encoding or decoding to be encoded or decoded using the constructed 8-bit primary transform core matrix that has the first size type that is 64-point or larger; and
when a result of the determination is to use the second transform core matrix of the second size type that is smaller than the first size type: causing or transmitting information that causes the target file for encoding or decoding to be encoded or decoded using the second transform core matrix of the second type.

### Claim 2 (depends on 1)

The method according to claim 1, wherein the processing steps of claim 1 are performed by one or more processors, and the method further comprises: storing, by the one or more processors in a memory, the second transform core matrix in advance before performing the processing steps of claim 1.

### Claim 3 (depends on 1)

The method according to claim 1, wherein
the first transform core matrix is a 64-point, 8-bit primary transform core matrix that is constructed using thirty-three integers which are used to construct a 32-point, 8-bit primary transform core matrix and an additional thirty-two integers that include: bf, bg, bh, bi, bj, bk, bl, bm, bn, bo, bp, bq, br, bs, bt, bu, by, bw, bx, by, bz, ca, cb, cc, cd, ce, cf, cg, ch, ci, cj and ck plus their respective sign changes, and
the additional thirty-two integers {bf, bg, bh, bi, bj, bk, bl, bm, bn, bo, bp, bq, br, bs, bt, bu, by, bw, bx, by, bz, ca, cb, cc, cd, ce, cf, cg, ch, ci, cj, ck} are equal to {91, 90, 90, 90, 88, 87, 86, 84, 83, 81, 79, 77, 73, 71, 69, 65, 62, 59, 56, 52, 48, 44, 41, 37, 33, 28, 24, 20, 15, 11, 7, 2}.

### Claim 4 (depends on 1)

The method according to claim 1, wherein the 128-point 8-bit primary transform core is constructed using the 65 integers used to construct a 64-point 8-bit primary transform core and additional 64 integers including: cl, cm, cn, co, cp, cq, cr, cs, ct, cu, cv, cw, cx, cy, cz, da, db, dc, dd, de, df, dg, dh, di, dj, dk, dl, dm, dn, d_, dp, dq, dr, ds, dt, du, dv, dw, dx, dy, dz, ea, eb, ec, ed, ee, ef, eg, eh, ei, ej, ek, el, em, en, eo, ep, eq, er, es, et, eu, ev and ew plus their sign changes, and these additional 64 integers are {cl, cm, cn, co, cp, cq, cr, cs, ct, cu, cv, cw, cx, cy, cz, da, db, dc, dd, de, df, dg, dh, di, dj, dk, dl, dm, dn, d_, dp, dq, dr, ds, dt, du, dv, dw, dx, dy, dz, ea, eb, ec, ed, ee, ef, eg, eh, ei, ej, ek, el, em, en, eo, ep, eq, er, es, et, eu, ev, ew} are equal to {90, 90, 90, 90, 90, 90, 89, 89, 89, 88, 88, 87, 86, 86, 85, 84, 83, 82, 81, 80, 79, 78, 77, 76, 75, 73, 72, 71, 69, 68, 66, 65, 63, 62, 60, 58, 58, 55, 53, 51, 49, 48, 45, 44, 42, 40, 38, 36, 34, 32, 29, 27, 25, 23, 21, 19, 17, 15, 12, 10, 8, 6, 3, 1}.

### Claim 5 (depends on 1)

The method according to claim 1, wherein the size of the first transform core matrix is 64-point or 128-point.

### Claim 6 (depends on 1)

The method according to claim 1, wherein
a left half or a right half of even or odd rows of the first transform core matrix form a matrix which is similar to the second transform core matrix,
the first transform core matrix is one of: DCT-2, DCT-5, DCT-8, DST-1 or DST-7, and
the second transform core matrix is one of: DCT-2, DCT-5, DCT-8, DST-1 or DST-7.

### Claim 7 (depends on 6)

The method according to claim 6, wherein the first transform core matrix and the second transform core matrix are the same one of: DCT-2, DCT-5, DCT-8, DST-1 or DST-7.

### Claim 8 (depends on 7)

The method according to claim 7, wherein the first transform core matrix is an 8-bit, 64-point DCT-2 transform matrix and the second transform core matrix is an 8-bit, 32-point DCT-2 transform matrix.

### Claim 9 (depends on 1)

The method according to claim 1, wherein the first transform core matrix is an 8-bit, N-point primary transform core matrix, which is composed by M unique numbers, and which is derived by the following steps:
deriving an initial transform core matrix T1 by scaling an original transform core which is composed by floating-point numbers T0 using a scaling factor 2B−2×log 2(N) such that T1=T0×2B-2×log2(N), where B=8, and the row vectors of T0, which are basis vectors, are orthogonal to each other, and the norm of each row vector is 1;
deriving an adjusted transform core matrix T2 by adding offsets on a selected set of the M unique numbers which are used to construct the adjusted primary transform core matrix T2;
calculating a cost value C as the sum of absolute values of all elements of (T1*T1T−22B-4×log2(N)2×I), where I is an N×N identity matrix; and
updating the adjusted transform matrix T2 by adding the calculated offset values to T2, which minimizes the cost value to T2, which is output as the 8-bit, N-point primary transform core matrix.

### Claim 10 (depends on 9)

The method according to claim 9, further comprising: after minimizing the cost value C by trying all possible combinations of offset values on a selected set of M unique numbers, T2 is output as the 8-bit, N-point primary transform core matrix.

### Claim 11

An apparatus for utilizing an 8-bit primary transform core matrix to decode compressed video or image data or encode uncompressed video or image data, the apparatus comprising:
at least one memory configured to store computer program code; and
at least one processor configured to access the at least one memory and operate according to the computer program code, the computer program code comprising:
first applying code configured to cause the at least one processor to receive information regarding a target file for encoding or decoding, the target file for encoding or decoding being one of: a compressed video or image data file or an uncompressed video or image data file;
second applying code configured to cause the at least one processor to determine, for the encoding or decoding of the target file, whether to use a first transform core matrix that is a Versatile Video Coding (VVC) standard transform and that has a first size type or use a second transform core matrix that is of a second size type that is smaller than the first size type, wherein the first size type is the 8-bit primary transform core matrix, which is represented by 8-bits and that has a size that is 64-point or larger;
third applying code configured to cause the at least one processor to when a result of the determination is to use the first transform core matrix of the first size type, construct the first transform core matrix using integers used to construct the second transform core matrix and additional integers, and cause or transmit information that causes the target file for encoding or decoding to be encoded or decoded using the constructed 8-bit primary transform core matrix that has the size that is 64-point or larger; and
fourth applying code configured to cause the at least one processor to when a result of the determination is to use the second transform core matrix of the second size type that is smaller than the first size type: cause or transmit information that causes the target file for encoding or decoding to be encoded or decoded using the second transform core matrix of the second type.

### Claim 12 (depends on 11)

The apparatus according to claim 11, wherein
the first transform core matrix is a 64-point, 8-bit primary transform core matrix that is constructed using thirty-three integers which are used to construct a 32-point, 8-bit primary transform core matrix and an additional thirty-two integers that include: bf, bg, bh, bi, bj, bk, bl, bm, bn, bo, bp, bq, br, bs, bt, bu, by, bw, bx, by, bz, ca, cb, cc, cd, ce, cf, cg, ch, ci, cj and ck plus their respective sign changes, and
the additional thirty-two integers {bf, bg, bh, bi, bj, bk, bl, bm, bn, bo, bp, bq, br, bs, bt, bu, by, bw, bx, by, bz, ca, cb, cc, cd, ce, cf, cg, ch, ci, cj, ck} are equal to {91, 90, 90, 90, 88, 87, 86, 84, 83, 81, 79, 77, 73, 71, 69, 65, 62, 59, 56, 52, 48, 44, 41, 37, 33, 28, 24, 20, 15, 11, 7, 2}.

### Claim 13 (depends on 11)

The apparatus according to claim 11, wherein the 128-point 8-bit primary transform core is constructed using the 65 integers used to construct a 64-point 8-bit primary transform core and additional 64 integers including: cl, cm, cn, co, cp, cq, cr, cs, ct, cu, cv, cw, cx, cy, cz, da, db, dc, dd, de, df, dg, dh, di, dj, dk, dl, dm, dn, d_, dp, dq, dr, ds, dt, du, dv, dw, dx, dy, dz, ea, eb, ec, ed, ee, ef, eg, eh, ei, ej, ek, el, em, en, eo, ep, eq, er, es, et, eu, ev and ew plus their sign changes, and these additional 64 integers are {cl, cm, cn, co, cp, cq, cr, cs, ct, cu, cv, cw, cx, cy, cz, da, db, dc, dd, de, df, dg, dh, di, dj, dk, dl, dm, dn, d_, dp, dq, dr, ds, dt, du, dv, dw, dx, dy, dz, ea, eb, ec, ed, ee, ef, eg, eh, ei, ej, ek, el, em, en, eo, ep, eq, er, es, et, eu, ev, ew} are equal to {90, 90, 90, 90, 90, 90, 89, 89, 89, 88, 88, 87, 86, 86, 85, 84, 83, 82, 81, 80, 79, 78, 77, 76, 75, 73, 72, 71, 69, 68, 66, 65, 63, 62, 60, 58, 58, 55, 53, 51, 49, 48, 45, 44, 42, 40, 38, 36, 34, 32, 29, 27, 25, 23, 21, 19, 17, 15, 12, 10, 8, 6, 3, 1}.

### Claim 14 (depends on 11)

The apparatus according to claim 11, wherein the size of the first transform core matrix is 64-point or 128-point.

### Claim 15 (depends on 11)

The apparatus according to claim 11, wherein
a left half or a right half of even or odd rows of the first transform core matrix form a matrix which is similar to the second transform core matrix,
the first transform core matrix is one of: DCT-2, DCT-5, DCT-8, DST-1 or DST-7, and
the second transform core matrix is one of: DCT-2, DCT-5, DCT-8, DST-1 or DST-7.

### Claim 16 (depends on 15)

The apparatus according to claim 15, wherein the first transform core matrix and the second transform core matrix are the same one of: DCT-2, DCT-5, DCT-8, DST-1 or DST-7.

### Claim 17 (depends on 16)

The apparatus according to claim 16, wherein the first transform core matrix is an 8-bit, 64-point DCT-2 transform matrix and the second transform core matrix is an 8-bit, 32-point DCT-2 transform matrix.

### Claim 18 (depends on 11)

The apparatus according to claim 11, wherein
the first transform core matrix is an 8-bit, N-point primary transform core matrix, which is composed of M unique numbers, and
the fourth applying code is further configured to cause the at least one processor to derive the 8-bit, N-point primary transform core matrix by performing the following steps:
deriving an initial transform core matrix T1 by scaling an original transform core which is composed by floating-point numbers T0 using a scaling factor 2B−2×log 2(N) such that T1=T0×2B-2×log2(N), where B=8, and the row vectors of T0, which are basis vectors, are orthogonal to each other, and the norm of each row vector is 1;
deriving an adjusted transform core matrix T2 by adding offsets on a selected set of the M unique numbers which are used to construct the adjusted primary transform core matrix T2;
calculating a cost value C as the sum of absolute values of all elements of (T1*T1T−22B-4×log2(N)2×I), where I is an N×N identity matrix; and
updating the adjusted transform matrix T2 by adding the calculated offset values to T2, which minimizes the cost value to T2, which is output as the 8-bit, N-point primary transform core matrix.

### Claim 19 (depends on 18)

The apparatus according to claim 18, wherein the fourth applying code is further configured to cause the at least one processor to: after minimizing the cost value C by trying all possible combinations of offset values on a selected set of the M unique numbers, cause T2 to be output as the 8-bit, N-point primary transform core matrix.

### Claim 20

A non-transitory computer-readable storage medium storing instructions that cause one or more processors to:
receive information regarding a target file for encoding or decoding, the target file for encoding or decoding being one of: a compressed video or image data file or an uncompressed video or image data file;
determine, for the encoding or decoding of the target file, whether to use a first transform core matrix that is a Versatile Video Coding (VVC) standard transform and that has a first size type or use a second transform core matrix that is of a second size type that is smaller than the first size type, wherein the first size type is the 8-bit primary transform core matrix, which is represented by 8-bits and that has a size that is 64-point or larger;
when a result of the determination is to use the first transform core matrix of the first size type, construct the first transform core matrix using integers used to construct the second transform core matrix and additional integers, and cause or transmit information that causes the target file for encoding or decoding to be encoded or decoded using the constructed 8-bit primary transform core matrix that has the first size type that is 64-point or larger; and
when a result of the determination is to use the second transform core matrix of the second size type that is smaller than the first size type: cause or transmit information that causes the target file for encoding or decoding to be encoded or decoded using the second transform core matrix of the second type.
