Video Coding Frame Loss Recovery via Dynamic Reference Hierarchy

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing video coding technologies face challenges in efficiently recovering lost frames, particularly in base layer frames, due to linear hierarchy structures and network conditions, leading to increased latency and bandwidth consumption.

Innovation Solution

The method dynamically updates the frame reference hierarchy by transmitting acknowledgements from the receiver to the sender, allowing the use of the last acknowledged base layer frame as a reference for subsequent frames, thereby minimizing frame losses and improving video smoothness without causing further latency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If traditional linear hierarchy reference frame structure is used, then encoding simplicity is maintained, but frame loss recovery efficiency deteriorates

Engineering Contradiction:

Improvereference frame structureVSAvoidframe loss recovery

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

The patent dynamically updates the reference frame structure based on acknowledgment feedback. When a base layer frame is lost, the system switches from a linear hierarchy to a dynamic hierarchy where subsequent frames reference the last successfully decoded base layer frame instead, enabling adaptive recovery without complex pre-planned structures

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system implements an acknowledgment mechanism where the receiver sends feedback to the sender about successfully decoded base layer frames. This feedback drives the dynamic updating of reference frame structures, allowing the system to adapt to actual transmission conditions and recover from losses efficiently

Inventive Principle:

Principle #23Feedback

2Reliability

If frame loss recovery mechanisms are implemented, then video smoothness is improved, but latency increases

Engineering Contradiction:

Improvevideo smoothnessVSAvoidlatency

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system proactively updates reference frame structures at the encoder side as soon as frame loss is detected through missing acknowledgments. This preliminary action allows subsequent frames to be encoded with correct references before playback, avoiding the need for reactive retransmission and reducing latency

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

When base layer frame loss is detected, the system skips waiting for retransmission and immediately switches to using the last acknowledged base layer frame as reference. This rushing through the recovery process maintains video smoothness while minimizing the time loss, preventing video freezes

Inventive Principle:

Principle #21Skipping (Rushing through)

3Reliability

If key frame retransmission is used for frame loss recovery, then frame loss recovery is achieved, but bandwidth consumption increases

Engineering Contradiction:

Improveframe loss recoveryVSAvoidbandwidth consumption

Core Design Contradiction:

ReliabilityVSLoss of energy

Solution Approach 1:

The patent extracts and utilizes intermediate base layer frames that were successfully decoded and stored at the receiver. Instead of retransmitting full key frames, the system reuses these existing intermediate frames as references, significantly reducing the bandwidth required for recovery while maintaining effective frame loss recovery

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS12108029B2Method and system of video coding with efficient frame loss recovery

Publication Date: 2024.10.01 INTEL CORP

AI summary

Techniques related to video coding include efficient frame loss recovery using a feedback channel acknowledgement to decode base layer frames.