Reference Picture Set Coding for Discardable Video Frames

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional video coding standards do not effectively prevent decoders from incorrectly inferring a loss when a discardable picture is dropped or incorrectly decoded from the bitstream, as they allow discardable pictures to be included in reference picture sets (RPS), leading to potential bandwidth inefficiencies and decoding errors.

Innovation Solution

The method involves determining whether a current picture is discardable and refraining from including it in the RPS, ensuring that discardable pictures are not used for inter-layer or temporal prediction, thereby preventing incorrect inference of losses during decoding.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If discardable pictures are included in reference picture sets (RPS), then the RPS contains more reference pictures for inter prediction, but decoders may incorrectly infer losses when discardable pictures are dropped or incorrectly decoded

Engineering Contradiction:
Improvereference picture availabilityVSAvoiddecoding accuracy
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The patent extracts discardable pictures from the reference picture set by checking the discardable picture flag before adding pictures to the RPS. This separation ensures that only non-discardable pictures are included in the reference picture lists used for inter prediction, preventing decoding errors while maintaining reliable reference picture availability.

Inventive Principle:
Principle #2Taking out (Extraction)

2Adaptability or versatility

If discardable pictures are included in RPS, then more reference pictures are available for inter-layer and temporal prediction, but bandwidth inefficiencies occur when these pictures are dropped from the bitstream

Engineering Contradiction:
Improveprediction capabilityVSAvoidbandwidth efficiency
Core Design Contradiction:
Adaptability or versatilityVSLoss of energy

Solution Approach 1:

The patent performs preliminary action by checking the discardable picture flag during the picture addition process to the RPS. This pre-filtering prevents discardable pictures from being included in the reference picture sets, ensuring that only pictures that will actually be transmitted and decoded are used for prediction, thereby avoiding bandwidth waste while maintaining prediction capability.

Inventive Principle:
Principle #10Preliminary action

3Productivity

If all decoded pictures are added to RPS without checking discardable flag, then the RPS is automatically populated, but decoding errors occur when discardable pictures are incorrectly inferred as lost

Engineering Contradiction:
Improveencoding speedVSAvoiddecoding correctness
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent implements feedback by continuously checking the discardable picture flag status before adding pictures to the RPS. This feedback mechanism ensures that the encoder and decoder maintain consistent understanding of which pictures are reference pictures and which are discardable, preventing decoding errors while maintaining efficient encoding throughput.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3090545A1Method for coding a reference picture set (RPS) in multi-layer coding
Publication Date: 2016.11.09 QUALCOMM INC

AI summary

A method for coding a reference picture set (RPS) in multi-layer coding is disclosed. In one aspect, the method may involve determining whether a current picture of video information is a discardable picture. The method may also involve refraining from including the current picture in an RPS based on the determination that the current picture is a discardable picture. The method may further involve encoding the video information based at least in part on the RPS.