EP-4736428-A1 - INTRA SUB-PARTITIONS WITH MULTIPLE PREDICTION MODES

Abstract

An apparatus may apply multiple coding modes to intra-prediction sub-partitions. The apparatus may apply intra-prediction to a coding block and may determine a first prediction mode associated with the coding block. The apparatus may determine at least a first sub-partition and a second partition associated with the coding block and may determine a second prediction mode that may be associated with at least one of the first sub-partition and the second sub-partition. The second prediction mode may comprise one of a mode in a list of most probable modes (MPM). The apparatus may determine a second prediction mode by decoding the first sub-partition using each of a plurality of prediction modes and selecting one of the plurality of prediction modes. The apparatus may generate decoded pixels associated with the first sub-partition and may copy the decoded pixels into the second sub-partition using a shift determined by the first prediction mode.

Inventors

• RATH, Gagan Bihari
• LE LEANNEC, FABRICE
• ROBERT, ANTOINE
• RADOSAVLJEVIC, Milos

Assignees

• InterDigital CE Patent Holdings, SAS

Dates

Publication Date

20260506

Application Date

20240620

Claims (15)

1. 1 . A device for video decoding comprising: a processor configured to: determine intra-prediction applies to a coding block; determine a first prediction mode associated with the coding block; determine a first sub-partition and a second sub-partition associated with the coding block; determine a second prediction mode, the second prediction mode associated with at least one of the first sub-partition or the second sub-partition, the second prediction mode being different than the first prediction mode; predict a first prediction for the first sub-partition using the first prediction mode; reconstruct the first sub-partition by adding the first prediction to decoded residual values for the first subpartition; predict a second prediction for the second sub-partition using the second prediction mode and based on the first sub-partition; and reconstruct the second sub-partition by adding the second prediction to decoded residual values for the second sub-partition.
2. 2. A device for video encoding comprising: determine intra-prediction applies to a coding block; determine a first prediction mode associated with the coding block; determine a first sub-partition and a second sub-partition associated with the coding block; determine a second prediction mode, the second prediction mode associated with at least one of the first sub-partition or the second sub-partition, the second prediction mode being different than the first prediction mode; predict a first prediction for the first sub-partition using the first prediction mode; reconstruct the first sub-partition by adding the first prediction to decoded residual values for the first sub-partition; predict a second prediction for the second sub-partition using the second prediction mode and based on the first sub-partition; and reconstruct the second sub-partition by adding the second prediction to decoded residual values for the second sub-partition.
3. 3. The device of claim 1 or 2, wherein the processor configured to predict the second prediction for the second sub-partition using the second prediction mode and based on the first sub-partition is further configured to predict the second prediction for the second sub-partition using the second prediction mode and based on a predicted sub-partition or a decoded sub-partition.
4. 4. The device of claim 1 or 2, wherein the second prediction mode comprises: a mode in a list of most probable modes (MPM), a mode equal to a prediction mode associated with a block adjacent to the first subpartition or the second sub-partition, or a fixed prediction mode.
5. 5. The device of claim 4, wherein the first prediction mode is not in the list of MPM; and wherein the second prediction mode is a first angular mode in the list of MPM.
6. 6. The device of claim 4, wherein the first prediction mode is a first angular mode in the list of MPM; and wherein the second prediction mode is a second angular mode in the list of MPM.
7. 7. The device of claim 1 or 2, wherein the processor configured to determine the second prediction mode is further configured to: decode the first sub-partition using a plurality of prediction modes; and select one of the plurality of prediction modes as the second prediction mode.
8. 8. A method of video decoding comprising: determining intra-prediction applies to a coding block; determining a first prediction mode associated with the coding block; determining a first sub-partition and a second sub-partition associated with the coding block; determining a second prediction mode, the second prediction mode associated with at least one of the first sub-partition or the second sub-partition, the second prediction mode being different than the first prediction mode; predicting a first prediction for the first sub-partition using the first prediction mode; reconstructing the first sub-partition by adding the first prediction to decoded residual values for the first subpartition; predicting a second prediction for the second sub-partition using the second prediction mode and based on the first sub-partition; and reconstructing the second sub-partition by adding the second prediction to decoded residual values for the second sub-partition.
9. 9. A method of video encoding comprising: determining intra-prediction applies to a coding block; determining a first prediction mode associated with the coding block; determining a first sub-partition and a second sub-partition associated with the coding block; determining a second prediction mode, the second prediction mode associated with at least one of the first sub-partition or the second sub-partition, the second prediction mode being different than the first prediction mode; predicting a first prediction for the first sub-partition using the first prediction mode; reconstructing the first sub-partition by adding the first prediction to decoded residual values for the first sub-partition; predicting a second prediction for the second sub-partition using the second prediction mode and based on the first sub-partition; and reconstructing the second sub-partition by adding the second prediction to decoded residual values for the second sub-partition.
10. 10. The method of claim 8 or 9, wherein predicting the second prediction for the second sub-partition using the second prediction mode and based on the first sub-partition further comprises predicting the second prediction for the second sub-partition using the second prediction mode and based on a predicted subpartition or a decoded sub-partition.
11. 11 . The method of claim 8 or 9, wherein the second prediction mode comprises: a mode in a list of most probable modes (MPM), a mode equal to a prediction mode associated with a block adjacent to the first subpartition or the second sub-partition, or a fixed prediction mode.
12. 12. The method of claim 11 , wherein the first prediction mode is not in the list of MPM; and wherein the second prediction mode is a first angular mode in the list of MPM.
13. 13. The method of claim 11 , wherein the first prediction mode is a first angular mode in the list of MPM; and wherein the second prediction mode is a second angular mode in the list of MPM.
14. 14. The method of claim 8 or 9, wherein determining the second prediction mode comprises: decoding the first sub-partition using a plurality of prediction modes; and selecting one of the plurality of prediction modes as the second prediction mode.
15. 15. A computer program product stored on a computer readable medium and comprising program code instructions for implementing the steps of a method according to any of claims 8 through 14 when executed by a processor.

Description

INTRA SUB-PARTITIONS WITH MULTIPLE PREDICTION MODES CROSS-REFERENCE TO RELATED APPLICATIONS [0001] This application claims the benefit of European Patent Application Number 23306085.4, filed June 30, 2023, the contents of which are hereby incorporated by reference herein in their entirety. BACKGROUND [0002] Video coding systems may be used to compress digital video signals, e.g., to reduce the storage and/or transmission bandwidth needed for such signals. Video coding systems may include, for example, block-based, wavelet-based, and/or object-based systems. SUMMARY [0003] Systems, methods, and instrumentalities are disclosed for performing intra-prediction using subpartitions with multiple prediction modes. [0004] A device which may be, for example, a video decoder device and/or video encoder device, may be configured to determine that intra-prediction applies to a coding block and may determine a first prediction mode that is associated with the coding block. [0005] The device may determine a first sub-partition and a second sub-partition associated with the coding block. The device may determine a second prediction mode. The second prediction mode may be associated with at least one of the first sub-partition or the second sub-partition and the second prediction mode may be different than the first prediction mode. The second prediction mode may be a mode in a list of most probable modes (MPM), a mode equal to a prediction mode associated with a block adjacent to the first sub-partition or second sub-partition, or a fixed prediction mode. In examples, the first prediction mode may not be in the list of MPM, and the second prediction mode may be a first angular mode in the list of MPM. In examples, the first prediction mode may be a first angular mode in the list of MPM, and the second prediction mode may be a second angular mode in the list of MPM. In examples, the second prediction mode may be determined by decoding the first sub-partition using a plurality of prediction modes, and selecting one of the plurality of prediction modes as the second prediction mode. [0006] The device may predict a first prediction for the first sub-partition using the first prediction mode and may reconstruct the first sub-partition by adding the first prediction to decoded residual values for the first sub-partition. [0007] The device may predict a second prediction for the second sub-partition using the second prediction mode and based on the first sub-partition (e.g., the predicted or decoded first sub-partition). The device may reconstruct the second sub-partition by adding the second prediction to decoded residual values for the second sub-partition. [0008] An apparatus, which may be, for example, an encoder or decoder, may be configured to determine a coding block which may be referred to as a coding unit. The apparatus may determine to apply intraprediction to the coding block and may determine a first prediction mode associated with the coding block. [0009] The apparatus may determine at least a first sub-partition and a second partition associated with the coding block. The apparatus may determine a second prediction mode that may be associated with at least one of the first sub-partition and the second sub-partition. The second prediction mode may comprise one of a mode in a list of most probable modes (MPM), a mode equal to a prediction mode associated with a block adjacent to the first sub-partition or second sub-partition, a fixed prediction mode, a vertical mode, or a horizontal mode. If the first prediction mode is not in the list of MPM, the second prediction mode may be a first angular mode in the list of MPM. If the first prediction mode is a first angular mode in the list of MPM, the second prediction mode may be a second angular mode in the list of MPM. [0010] The apparatus may be configured to determine a second prediction mode by decoding the first sub-partition using each of a plurality of prediction modes. The plurality of prediction modes may comprise the modes designated in a list of MPM. The apparatus may select one of the plurality of prediction modes as the second prediction mode. The selected one of the plurality of prediction modes may be associated with the second sub-partition. [0011] The apparatus may predict the first sub-partition using the first prediction mode and may predict the second sub-partition using the second prediction mode. [0012] The apparatus may be configured to generate decoded pixels associated with the first sub-partition. The apparatus may copy the decoded pixels into the second sub-partition using a shift determined by the first prediction mode. The shift may be, for example, a horizontal shift or a vertical shift. [0013] Systems, methods, and instrumentalities described herein may involve a decoder. In some examples, the systems, methods, and instrumentalities described herein may involve an encoder. In some examples, the systems, methods, and instrumentalities descr