Video Sequence Reference Picture Set Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High Efficiency Video Coding (HEVC) requires efficient signaling of Reference Picture Sets (RPSs) to reduce bitrates, as existing methods are not optimized for minimizing bit usage, especially at low bit rates and small image sizes.
Innovation Solution
The proposed method arranges RPSs in transmission order within a Sequence Parameter Set (SPS) and uses inter-prediction to infer or explicitly encode RPS information only when necessary, reducing the need for explicit signaling and minimizing bit usage by omitting unnecessary parameters like delta_idx_minus1.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If RPS information is explicitly signaled in slice headers, then decoding accuracy is improved, but bitrate increases significantly
Solution Approach 1:
The patent applies preliminary action by signaling RPS information in the SPS (Sequence Parameter Set) before the actual video slices are encoded. By pre-defining the reference picture sets in the SPS, the decoder can reconstruct RPS information without requiring extensive explicit signaling in each slice header, thus reducing bitrate while maintaining decoding accuracy
Solution Approach 2:
The patent uses copying by reusing previously defined RPS configurations from the SPS for subsequent slices. Instead of explicitly signaling complete RPS information for every slice, the encoder copies reference picture set definitions from the SPS, significantly reducing the number of bits required while ensuring consistent decoding behavior
2Measurement precision
If complete RPS information is transmitted for every slice, then reference picture accuracy is improved, but data redundancy increases
Solution Approach 1:
The patent extracts the RPS information that is common across multiple slices and places it in the SPS, separating it from the slice-specific data. This extraction eliminates redundancy by ensuring that reference picture set definitions are transmitted only once in the SPS rather than being repeated in every slice header, while still maintaining reference picture accuracy through the use of delta values for slice-specific modifications
3Quantity of substance
If RPS is sent in SPS only, then bitrate is reduced, but flexibility for slice-specific reference pictures is lost
Solution Approach 1:
The patent applies local quality by allowing different slices to have different levels of RPS signaling detail. The SPS contains the base RPS configuration for all slices, while individual slices can use delta values to locally modify the reference picture sets as needed. This enables slice-specific reference picture flexibility while maintaining overall bitrate efficiency through the hierarchical signaling structure
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
A method (1000) of encoding a video sequence comprising Reference Picture Sets (RPSs) is provided. The method comprises arranging (1001) the RPSs in transmission order in a data structure, such as a Sequence Parameter Set (SPS), determining (1003) whether explicit RPS transmission is used for an RPS of a current picture of the video sequence, and encoding (1004) information indicating an RPS comprised in the data structure to be used for predicting the RPS of the current picture, such as delta_idx_minus1, only if explicit RPS transmission is used. By transmitting delta_idx_minus1 only if explicit RPS transmission is used, and interpreting delta_idx_minus1 to be equal to zero otherwise, a reduced bitrate is achieved. Further, a method of decoding a video sequence comprising RPSs, corresponding computer programs and computer program products, as well as corresponding encoders and decoders are provided.