Online and offline selection of extended long term reference picture retention

Patent metadata

Field	Value
Publication number	US-12470694-B2
Application number	US-202318380704-A
Country	US
Kind code	B2
Filing date	Oct 17, 2023
Priority date	Jan 28, 2019
Publication date	Nov 11, 2025
Grant date	Nov 11, 2025

Abstract

A video encoder encodes a plurality of pictures with one of the pictures being a long term reference picture. The encoder provides signaling in a bitstream for a decoder to reconstruct pictures by selectively activating the long term reference picture for use as a reference picture, maintaining the long term reference picture in the buffer while being inactive as a reference picture, changing the long term reference picture from inactive to active as a reference picture, and marking the long term reference picture as being unused for reference so that it is no longer eligible to be used as a reference picture for a subsequent picture and is removed as a reference picture.

First claim

What is claimed is:

1 . A video encoder for generating an encoded bitstream with signaling for a decoder, the encoder comprising:
a motion compensation processor; an inverse quantizer and inverse transformer, wherein a plurality of pictures being reconstructed by using the inverse transformer and inverse quantizer to inverse transform and inverse quantize residual signals, and adding the residual signals to predictors generated by the motion compensation processor, the plurality of reconstructed pictures including first, second, and third reconstructed pictures; and a buffer for storing the reconstructed pictures including the first, second and third pictures, wherein the reconstructing comprises:
reconstructing the first picture by determining a first list of active reference pictures from the pictures in the buffer, one of the active reference pictures being a long term reference picture, and reconstructing a block of said first picture using a predictor from the long term reference picture;
reconstructing the second picture by determining a second list of active reference pictures from the pictures in the buffer, said second list not including the long term reference picture and, in response to at least one parameter generated by an encoder, determining that the long term reference picture has a status where it is inactive as a reference picture for the second picture while continuing to be stored in the buffer and eligible to be an active reference picture for a third picture; and
reconstructing the third picture by determining a third list of active reference pictures from the pictures in the buffer including, in response to at least one parameter generated by an encoder, changing the determination of the long term reference picture from inactive to active as a reference picture, and including the long term reference picture in the third list, and reconstructing a block from the third picture using a predictor from the long term reference picture,
the long term reference picture then being marked as unused for reference so that it is not eligible to be used as a reference picture for a subsequent picture and is removed from the buffer.

2 . The encoder of claim 1 , wherein the plurality of pictures are single view pictures.

3 . The encoder of claim 1 , wherein the long term reference picture is removed from the buffer when an increasing counting parameter reaches a predetermined value.

4 . The encoder of claim 3 , wherein the increasing counting parameter is a time related parameter.

5 . A method of video signal processing comprising:
reconstructing a plurality of pictures, the plurality of reconstructed pictures including first, second, and third reconstructed pictures; and storing the first, second and third reconstructed pictures, wherein the reconstructing a plurality of pictures comprises:
reconstructing the first picture by determining a first list of active reference pictures, one of the active reference pictures being a long term reference picture, and reconstructing a block of said first picture using a predictor from the long term reference picture;
reconstructing the second picture by determining a second list of active reference pictures, said second list not including the long term reference picture and, in response to at least one parameter generated by an encoder, determining that the long term reference picture has a status where it is inactive as a reference picture for the second picture while continuing to be stored and eligible to be an active reference picture for a third picture; and
reconstructing the third picture by determining a third list of active reference pictures, in response to at least one parameter generated by an encoder changing the determination of the long term reference picture from inactive to active as a reference picture, and including the long term reference picture in the third list, and reconstructing a block from the third picture using a predictor from the long term reference picture; and
the long term reference picture then being marked as unused for reference so that it is ineligible to be used as a reference picture for a subsequent picture and is removed from storage.

6 . The encoder of claim 5 , wherein the long term reference picture is removed from the buffer when an increasing counting parameter reaches a predetermined value.

7 . The encoder of claim 6 , wherein the increasing counting parameter is a time related parameter.

Assignees

• Op Solutions Llc

Inventors

• Furht Borivoje
• Kalva Hari
• Adzic Velibor

Classifications

• H04N19/58 Primary

  Motion compensation with long-term prediction, i.e. the reference frame for a current frame not being the temporally closest one (H04N19/23 takes precedence) · CPC title

• H04N19/46

  Embedding additional information in the video signal during the compression process (H04N19/517, H04N19/68, H04N19/70 take precedence) · CPC title

• H04N19/172

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

• H04N19/70

  characterised by syntax aspects related to video coding, e.g. related to compression standards · CPC title

• H04N19/105 Primary

  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 · CPC title