Video Coding History-Based List for Intra Prediction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current video coding technologies face challenges in efficiently reducing redundancy in video signals, particularly in representing less likely intra prediction directions and motion vectors, which affects compression efficiency and bandwidth requirements.

Innovation Solution

The proposed solution involves a method for video encoding and decoding that uses a history-based list to store and update sample locations, allowing for the reconstruction of a string of sample values within a current block, and employs a single value string mode for flexible sample reconstruction, optimizing the representation of intra prediction directions and motion vectors.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If traditional intra coding techniques are used, then the coding process is simple, but the compression efficiency is insufficient for representing less likely prediction directions and motion vectors

Engineering Contradiction:
Improvecompression efficiencyVSAvoidcoding process complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent pre-calculates and stores cumulative frequency values for all possible intra prediction directions and motion vector differences during the encoding process. This preliminary action allows the decoder to directly reconstruct these values without complex calculations, thereby improving compression efficiency while maintaining manageable complexity through pre-computation.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent creates a copy of the reference block and uses it for generating prediction samples. By copying the reference block data and storing it in a buffer, the system can efficiently generate predictions for multiple directions without repeatedly accessing the original reference picture, thus improving compression efficiency for less likely prediction directions.

Inventive Principle:
Principle #26Copying

2Adaptability or versatility

If more intra prediction directions are supported, then the adaptability increases, but the bandwidth requirements increase

Engineering Contradiction:
Improveprediction direction coverageVSAvoidbandwidth requirements
Core Design Contradiction:
Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

The patent implements a dynamic coding strategy where the precision of motion vector difference representation adapts based on the intra prediction direction. For less likely prediction directions, the system uses higher precision (more bits) to accurately represent the motion vector differences, while for more likely directions, it uses lower precision. This dynamic approach increases adaptability across different prediction scenarios while optimizing bandwidth usage by allocating bits only where needed.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes the precision parameter of motion vector difference representation based on the intra prediction direction. By adjusting the number of bits used to represent motion vector differences according to the specific prediction direction being used, the system achieves better adaptability for various directions while controlling bandwidth requirements through selective precision adjustment.

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If a history-based list stores all sample locations, then the reconstruction accuracy improves, but the memory requirements increase

Engineering Contradiction:
Improvesample value reconstruction accuracyVSAvoidmemory buffer size
Core Design Contradiction:
Measurement precisionVSVolume of stationary object

Solution Approach 1:

The patent extracts only the necessary sample locations from the reference block and stores them in a buffer, rather than storing all possible sample locations. By selectively taking out and storing only the locations that are actually needed for prediction, the system achieves good reconstruction accuracy while minimizing memory buffer size requirements.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent uses a partial approach by storing a limited set of sample locations in the buffer rather than all possible locations. This partial action is sufficient to achieve good reconstruction accuracy for the predicted blocks while keeping the memory buffer size manageable. The buffer stores enough information to handle the prediction requirements without excessive memory consumption.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentUS11627328B2Method and apparatus for video coding
Publication Date: 2023.04.11 TENCENT AMERICA LLC
  • US11627328B2 patent drawing
  • US11627328B2 patent drawing
  • US11627328B2 patent drawing

AI summary

Aspects of the disclosure provide methods and apparatuses for video encoding/decoding. In some examples, an apparatus for video decoding includes processing circuitry. The processing circuitry determines an area with sample values available in a reference sample memory for use in a reconstruction of a current block. Further, the processing circuitry updates a history based list for storing locations of previously reconstructed samples of a single value string mode. The locations are limited within the area. Then, the processing circuitry reconstructs, based on the history based list, a string of the single value string mode within the current block.