Methods and apparatus to encode and decode video using quantization matrices

Patent metadata

Field	Value
Publication number	US-12574516-B2
Application number	US-202117555092-A
Country	US
Kind code	B2
Filing date	Dec 17, 2021
Priority date	Dec 17, 2021
Publication date	Mar 10, 2026
Grant date	Mar 10, 2026

Abstract

Methods, apparatus, systems, and articles of manufacture are disclosed to encode and decode video using quantization matrices. An example apparatus includes interface circuitry to access an input frame of video, quantization matrix syntax encoder circuitry to encode a set of user-defined quantization matrices into a sequence header associated with a sequence of video frames including the input frame, adaptive quantization matrix selector circuitry to select a subset of quantization matrices from a combination of a set of default quantization matrices and the set of user-defined quantization matrices, adaptive segment selector circuitry to select a first one of the subset of quantization matrices for a first segment of the input frame, the input frame to be divided into a plurality of segments including the first segment, and encoder circuitry to quantize transform coefficients of the first segment of the input frame based on the first one of the subset of quantization matrices.

First claim

What is claimed is:

1 . An apparatus to encode video, the apparatus comprising:
interface circuitry to access a first an initial input frame and a second subsequent input frame of video; machine readable instructions; and processor circuitry to at least one of instantiate or execute the machine readable instructions to:
encode a set of user-defined quantization matrices into a sequence header associated with a sequence of video frames including the initial input frame and the subsequent input frame;
quantize the initial input frame based on a first subset of quantization matrices from a combination of a set of default quantization matrices and the set of user-defined quantization matrices;
after the quantization of the initial input frame, select a different second subset of quantization matrices for the subsequent input frame based on encoding information from the initial input frame, the second subset of quantization matrices and the first subset of quantization matrices having at least one quantization matrix not in common, the second subset of quantization matrices from a combination of the set of default quantization matrices and the set of user-defined quantization matrices; and
quantize transform coefficients of the subsequent input frame based on the second subset of quantization matrices.

2 . The apparatus of claim 1 , including example picture analyzer circuitry to analyze spatial features of the initial input frame and the subsequent input frame.

3 . The apparatus of claim 1 , wherein the processor circuitry is to;
encode the initial input frame based on the first subset of quantization matrices to determine an encoded frame; and select the second subset of quantization matrices based on one or more of an error rate or quality parameter associated with the encoded frame.

4 . The apparatus of claim 1 , wherein the set of user-defined quantization matrices is derived based on at least one machine learning technique, ones of the set of user-defined quantization matrices to be associated with corresponding different video characteristics.

5 . The apparatus of claim 1 , wherein a number of quantization matrices in the second subset equals a number of default quantization matrices, and to select the second subset of quantization matrices, the processor circuitry is to select ones of the user-defined quantization matrices to replace corresponding ones of the default quantization matrices in the second subset of quantization matrices.

6 . The apparatus of claim 1 , wherein a number of quantization matrices in the second subset is greater than a number of default quantization matrices, and to select the second subset of quantization matrices, the processor circuitry is to select ones of the user-defined quantization matrices in addition to ones of the default quantization matrices to include in the second subset of quantization matrices.

7 . The apparatus of claim 1 , wherein the set of user-defined quantization matrices is empty, and to select the second subset of quantization matrices, the processor circuitry is to implicitly select quantization matrices based on a quantization parameter.

8 . The apparatus of claim 7 , wherein the quantization parameter is one of multiple quantization parameters, and to implicitly select quantization matrices based on a quantization parameter, the processor circuitry is to:
classify the multiple quantization parameters into categories; and map a first one of the categories to a first one of the set of default quantization matrices based on values of the quantization parameters in the first one of the categories and flatness of the first one of the set of default quantization matrices.

9 . The apparatus of claim 1 , wherein the processor circuitry is to encode the selection of the second subset of quantization matrices into a frame header, a number of bits used to encode the selection to be based on a total number of quantization matrices within a combination of the set of default quantization matrices and the set of user-defined quantization matrices.

10 . The apparatus of claim 1 , wherein:
the subsequent input frame is divided into a plurality of segments including a first segment; and the processor circuitry is to:
select a first one of the second subset of quantization matrices to be used to quantize the first segment; and
encode the selection of the first one of the second subset of quantization matrices into a frame header, a number of bits used to encode the selection to be based on a number of quantization matrices within the subset of quantization matrices.

11 . At least one non-transitory machine-readable medium comprising instructions to cause at least one processor circuit to at least:
access an initial input frame and a subsequent input frame of video; encode a set of user-defined quantization matrices into a sequence header associated with a sequence of video frames including the initial input frame and the subsequent input frame; quantize the initial input frame based on a first subset of quantization matrices from a combination of a set of default quantization matrices and the set of user-defined quantization matrices; after the quantization of the initial input frame, select a different second subset of quantization matrices for the subsequent input frame based on encoding information from the initial input frame, the second subset of quantization matrices and the first subset of quantization matrices having at least one quantization matrix not in common, the second subset of quantization matrices from a combination of the set of default quantization matrices and the set of user-defined quantization matrices; and quantize transform coefficients of the subsequent input frame based on the second subset of quantization matrices.

12 . The at least one non-transitory machine-readable medium of claim 11 , wherein the instructions are to cause one or more of the at least one processor circuit to analyze spatial features of the initial input frame and the subsequent input frame.

13 . The at least one non-transitory machine-readable medium of claim 12 , wherein the instructions are to cause one or more of the at least one processor circuit to;
encode the initial input frame of the video based on the first subset of quantization matrices to determine an encoded frame; and select the second subset of quantization matrices based on one or more of an error rate or quality parameter associated with the encoded frame.

14 . The at least one non-transitory machine-readable medium of claim 11 , wherein the set of user-defined quantization matrices is derived based on at least one machine learning technique, the set of user-defined quantization matrices to include quantization matrices for various videos.

15 . The at least one non-transitory machine-readable medium of claim 11 , wherein a number of quantization matrices in the second subset equals a number of default quantization matrices, and to select the second subset of quantization matrices, the instructions are to cause one or more of the at least one processor circuit to select ones of the user-defined quantization matrices to replace corresponding ones of the default quantization matrices in the second subset of quantization matrices.

16 . The at least one non-transitory machine-readable medium of claim 11 , wherein a number of quantization matrices is greater than a number of default quantization matrices, and to select the second subset of quantization matrices, the instructions are to cause one or more of the at least one processor circuit to selec

Assignees

• Intel Corp

Inventors

• Zhang Ximin
• Lei Zhijun
• Boyce Jill
• Lee Sang-Hee

Classifications

• H04N19/119

  Adaptive subdivision aspects, e.g. subdivision of a picture into rectangular or non-rectangular coding blocks · CPC title

• H04N19/18

  the unit being a set of transform coefficients · CPC title

• H04N19/172

  the region being a picture, frame or field · CPC title

• H04N19/136

  Incoming video signal characteristics or properties · CPC title

• H04N19/423

  characterised by memory arrangements (H04N19/433 takes precedence) · CPC title