Cross-Retargeted Inter Prediction for Mixed-ROI Video Decoding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional video coding systems face challenges in effectively integrating inter prediction with retargeting, particularly when dealing with video sequences containing varying proportions and resolutions of Regions of Interest (ROIs), leading to inefficiencies and suboptimal performance.

Innovation Solution

The implementation of cross-retargeting, where a first encoded picture is adapted to comply with the retargeting scheme of a second encoded picture, allowing for consistent inter prediction across varying retargeting schemes without impacting the fidelity of the retargeting process.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If retargeting is applied to enhance fidelity for regions of interest with adaptive scaling, then fidelity for ROIs is improved, but consistency between different encoded pictures is deteriorated

Engineering Contradiction:
ImprovefidelityVSAvoidconsistency
Core Design Contradiction:
Measurement precisionVSStability of the object's composition

Solution Approach 1:

The patent introduces a reference picture adaptation mechanism that acts as an intermediary between pictures with different retargeting schemes. The decoder adapts the reference picture to match the retargeting scheme of the current picture being decoded, enabling effective inter prediction while preserving ROI fidelity. This mediator resolves the conflict between individual picture optimization and sequence-wide consistency.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent implements dynamic retargeting scheme adaptation where the reference picture's retargeting parameters are adjusted based on the current picture's requirements. Instead of using a fixed retargeting approach, the system dynamically modifies scaling factors and ROI configurations to match each picture's specific retargeting scheme, enabling both high fidelity and consistency.

Inventive Principle:
Principle #15Dynamics

2Productivity

If inter prediction is used to achieve higher compression rates, then compression efficiency is improved, but effectiveness with varying retargeting schemes is deteriorated

Engineering Contradiction:
Improvecompression efficiencyVSAvoideffectiveness with varying retargeting schemes
Core Design Contradiction:
ProductivityVSAdaptability or versatility

Solution Approach 1:

The patent changes the parameters of the reference picture (scaling factors, ROI positions, retargeting scheme) to match the current picture being decoded. By dynamically adjusting these parameters based on the picture pair's specific retargeting configurations, the system maintains effective inter prediction across varying retargeting schemes while preserving compression efficiency.

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If retargeting adapts each picture independently to its own scheme, then fidelity for that picture is improved, but compatibility for inter prediction is deteriorated

Engineering Contradiction:
ImprovefidelityVSAvoidcompatibility
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

Instead of trying to make each picture adapt to a common retargeting scheme (the obvious approach), the patent inverts the problem by adapting the reference picture to match the current picture's retargeting scheme. This inversion allows each picture to maintain its optimized retargeting configuration while still enabling compatible inter prediction through the adapted reference picture.

Inventive Principle:
Principle #13The other way round (Inversion)

Data Source

PatentUS20260025514A1Cross-retargeting for inter prediction
Publication Date: 2026.01.22 GUANGDONG OPPO MOBILE TELECOMMUNICATIONS CORP LTD
  • US20260025514A1 patent drawing
  • US20260025514A1 patent drawing
  • US20260025514A1 patent drawing

AI summary

A method of processing video data, performed by a decoder, is provided. The method includes that: a bitstream including a first picture retargeted according to a first retargeting scheme and a second picture retargeted according to a second retargeting scheme is received; the first picture is decoded from the bitstream to generate a first decoded picture; the first decoded picture is adapted to the second retargeting scheme to generate a reference picture; the second picture is decoded from the bitstream to generate a second decoded picture, wherein the operation of decoding the second picture includes inter prediction using the reference picture; and an output sequence including the first and second decoded pictures is generated.