Transcoder Using Pre-computed Frame Information for Video Representation Switching

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current video delivery systems face challenges in providing flexible switching between different video representations with reduced computational complexity, especially when dealing with limited processing power at intermediate stages and varying network bandwidths.

Innovation Solution

A transcoder that receives a basic video representation and frame information sets, decodes, and re-encodes on a frame-by-frame basis using provided coding information to generate arbitrary representations, reducing the need for complex encoding decisions and allowing for flexible generation of desired video streams with varying technical characteristics.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If full encoding is performed at intermediate transcoders to enable flexible switching between different video representations, then adaptability is improved, but device complexity and computational power requirements worsen

Engineering Contradiction:

Improveflexibility to switch between different video representationsVSAvoidcomputational complexity of transcoder

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The encoder at the source side performs encoding decisions in advance for multiple possible representations (different bit rates, resolutions, frame rates). These pre-computed encoding decisions are packaged in frame information packets and transmitted to intermediate transcoders, eliminating the need for complex real-time encoding decisions at intermediate stages.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The video stream is divided into segments with multiple representations. Each segment contains frame information packets that include pre-computed encoding decisions for different representations. Intermediate transcoders can selectively choose and switch between these pre-segmented representations without performing full encoding operations.

Inventive Principle:

Principle #1Segmentation

2Adaptability or versatility

If multiple full encoders are deployed at intermediate stages to generate arbitrary representations, then adaptability is improved, but use of energy and processing power worsen

Engineering Contradiction:

Improveability to generate arbitrary video representationsVSAvoidprocessing power consumption at intermediate stages

Core Design Contradiction:

Adaptability or versatilityVSUse of energy by moving object

Solution Approach 1:

Instead of deploying multiple full encoders at intermediate stages, the system creates copies of pre-computed frame information packets from the source encoder. These packets contain encoding decisions for multiple representations and can be selectively used at intermediate transcoders, avoiding the need for energy-intensive re-encoding operations.

Inventive Principle:

Principle #26Copying

Solution Approach 2:

All encoding decisions for multiple representations are computed in advance at the source side before transmission. Intermediate transcoders only need to perform lightweight selection and switching operations on pre-computed representations, dramatically reducing energy consumption compared to real-time full re-encoding.

Inventive Principle:

Principle #10Preliminary action

3Adaptability or versatility

If video streams are transmitted with multiple representations at the source side, then adaptability is improved, but loss of information and bandwidth requirements worsen

Engineering Contradiction:

Improveflexibility to accommodate different network bandwidthsVSAvoiddata volume transmitted over the network

Core Design Contradiction:

Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

The system extracts only the essential frame information packets containing encoding decisions from the full video representations. These compact packets are transmitted over the network to intermediate transcoders, which then reconstruct the desired representations locally. This approach transmits minimal data while maintaining the ability to generate multiple representations.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

Encoding decisions for multiple representations are pre-computed and packaged in compact frame information packets at the source side. Only these lightweight packets are transmitted over the network, not the full video streams. Intermediate transcoders use these packets to efficiently generate the required representations, reducing network bandwidth requirements.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS11356683B2Creating different video stream representations

Publication Date: 2022.06.07 INTERUNIVERSITAIR MICRO ELECTRONICS CENT (IMEC VZW)

AI summary

A transcoder for transcoding a basic representation of a videostream into one or more arbitrary representations comprises: a receiver adapted for receiving the basic representation and at least one frame information set comprising one or more frame information packets; a decoder adapted for decoding the basic representation; a re-encoder adapted for selecting at least one frame information set and for selecting one or more frame information packets from this at least one frame information set for forming a arbitrary representation, the re-encoder is adapted for extracting coding information from the frame information packets, and for re-encoding the decoded basic representation using the coding information thereby obtaining the one or more arbitrary representations.