Reference Picture List Derivation for Zero-Entry Video Prediction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In video coding and decoding, the management of reference picture lists can result in an undetermined reference picture when the number of entries is 0, leading to inefficiencies and uncertainties in the prediction process.

Innovation Solution

A video decoding and coding apparatus that derives reference picture lists by using flags and structures included in sequence and picture headers, estimating the number of entries to be 0 when the first reference picture list is not derived, ensuring a defined reference picture is used for prediction image generation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If a reference picture list with 0 entries is defined in the reference picture list management system, then the system allows flexible configuration of reference picture lists, but the reference picture becomes undetermined

Engineering Contradiction:
Improvereference picture list configuration flexibilityVSAvoidreference picture determination
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The patent applies preliminary anti-action by introducing a flag mechanism that prevents the undetermined reference picture issue before it occurs. When the number of reference pictures is 0, the flag is set to indicate this state, and the system proactively selects a default reference picture (such as the most recent decoded picture) instead of allowing the reference picture to remain undetermined. This preemptive measure eliminates the reliability problem while preserving the flexibility to define lists with 0 entries.

Inventive Principle:
Principle #9Preliminary anti-action

2Adaptability or versatility

If multiple reference picture lists are defined and managed, then the prediction image generation can use more reference pictures, but the complexity of managing reference picture list structures increases

Engineering Contradiction:
Improvereference picture list managementVSAvoidreference picture list structure
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent applies self-service by enabling the decoding apparatus to automatically derive and manage reference picture lists without requiring explicit detailed configuration for each list. The system uses flags and compact syntax structures that allow the decoder to self-determine the number of reference pictures and select appropriate reference pictures based on predefined rules and the current decoding context, thereby reducing the complexity of managing multiple reference picture lists.

Inventive Principle:
Principle #25Self-service

3Productivity

If the number of reference pictures is set to 0 to optimize coding efficiency, then the coded data size can be reduced, but the prediction image generation cannot proceed without a determined reference picture

Engineering Contradiction:
Improvecoding efficiencyVSAvoidprediction image generation
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent applies preliminary action by pre-defining default reference picture selection rules that are established before the actual prediction image generation process. When the number of reference pictures is 0, the system has already predetermined which picture should be used as the reference (such as the most recently decoded picture or a picture from a default list), allowing the prediction image generation to proceed smoothly without requiring additional runtime determination, thus maintaining both coding efficiency and prediction reliability.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS12634442B2Video decoding apparatus, video coding apparatus, video decoding method, and video coding method
Publication Date: 2026.05.19 SHARP KK
  • US12634442B2 patent drawing
  • US12634442B2 patent drawing
  • US12634442B2 patent drawing

AI summary

A video decoding apparatus is a video decoding apparatus for deriving a prediction image by using a reference picture included in a reference picture list. The video decoding apparatus includes: a parameter decoder configured to decode (i) one or more reference picture list structures, and (ii) a first flag indicating whether first reference picture list information is present in a slice header whose nal_unit_type is an IDR picture, the one or more reference picture list structures and the first flag being included in a sequence parameter set, and a second flag indicating whether or not second reference picture list information is present in a picture header, the second flag being included in a picture parameter set; and a prediction parameter derivation unit configured to (i) derive the first reference picture list information by using nal_unit_type or the first flag and the second flag in the slice header, or derive the second reference picture list information by using the second flag in the picture header, and (ii) derive the reference picture list, based on the one or more reference picture list structures. In a case that the first reference picture list is not derived, the number of entries of the one or more reference picture list structures is estimated to be 0.