3D Merge List Size Adjustment for Video Coding Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 3D-HEVC standards do not effectively utilize the view synthesis prediction flag to adjust the size of the Merge candidate list, limiting coding performance and flexibility.
Innovation Solution
The proposed method determines a modified list size for the 3D Merge list by considering the view synthesis prediction flag, along with other 3D enabling flags, to include additional candidates, thereby enhancing the Merge candidate list size and improving coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the Merge candidate list size is increased to include additional 3D Merge candidates (IVMP, DV, MPI, DDD, VSP), then coding performance and flexibility are improved, but the complexity of deriving and managing the candidate list increases
Solution Approach 1:
The patent implements dynamic adjustment of the Merge candidate list size based on the view synthesis prediction flag. When VSP is enabled, the list size is increased to include additional 3D Merge candidates (IVMP, DV, MPI, DDD, VSP). This dynamic approach allows the system to adapt the candidate list configuration to the specific coding scenario, improving coding efficiency when 3D prediction is beneficial while avoiding unnecessary complexity when it is not needed.
2Adaptability or versatility
If the view synthesis prediction flag is utilized to adjust the Merge candidate list size, then design flexibility and coding performance are enhanced, but the requirement for additional flag signaling and processing increases
Solution Approach 1:
The view synthesis prediction flag serves multiple functions in the patent. It not only enables or disables the VSP feature but also dynamically controls the size and composition of the Merge candidate list. This multi-functionality allows a single flag to provide both feature control and candidate list configuration, enhancing design flexibility while minimizing the need for additional separate signaling mechanisms.
3Productivity
If additional 3D Merge candidates (IVMP, DV, MPI, DDD, VSP) are included in the Merge candidate list, then compression efficiency for dependent-view textures and depth maps is improved, but the computational resources required for candidate evaluation increase
Solution Approach 1:
The patent applies local quality by selectively including different types of 3D Merge candidates in the candidate list based on the specific coding context and the enabled prediction modes. Not all five types of additional candidates (IVMP, DV, MPI, DDD, VSP) are included in every case - only those relevant to the current coding scenario are added. This selective approach improves compression efficiency for dependent-view textures and depth maps while avoiding the computational overhead of evaluating unnecessary candidates.
Data Source
AI summary
A method and apparatus for 3D (three-dimensional) and multi-view video encoding or decoding of dependent-view texture or depth data using Merge mode with modified list size according to one or more 3D enabling flags are disclosed. An extra-candidate-number indicating one or more extra Merge candidates for the current block is determined according to one or more 3D enabling flags comprising a VSP (view synthesis prediction) flag. A modified list size for a 3D Merge list corresponding to a sum of the extra-candidate-number and an original list size is determined. The 3D Merge list with the modified list size is then determined based on original Merge candidates used to derive the base-view Merge list for video data in a base view and one or more 3D Merge candidates as enabled by the 3D enabling flags according to a priority order.