Video Prediction Signaling for Adaptive Resolution Changes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding standards lack the ability to adaptively change resolution without introducing intra random access points, which can lead to increased decoding complexity, resource consumption, and poor user experience in applications like video conferencing and streaming.
Innovation Solution
Implementing adaptive resolution change (ARC) and reference picture resampling (RPR) techniques to allow for seamless resolution adjustments, including the use of Decoder-side Motion Vector Refinement and Bi-Directional Optical Flow coding tools, and defining conformance windows to manage resolution changes efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If adaptive resolution change and reference picture resampling are not implemented, then decoding complexity remains low, but coding performance deteriorates due to inaccurate motion information and improper handling of resolution differences
Solution Approach 1:
The patent changes the resolution parameter of reference pictures through resampling operations. By adjusting the resolution parameters to match between current and reference pictures, the system achieves accurate motion compensation without requiring complex decoding procedures. The resampling process modifies the spatial dimensions of reference pictures to align with current picture resolution, thereby improving coding performance while maintaining manageable decoding complexity.
Solution Approach 2:
The patent introduces an intermediate resampling process that acts as a mediator between reference pictures and motion compensation operations. This intermediary step resamples reference pictures to match the resolution of current pictures, enabling accurate motion vector application without directly complicating the decoding process. The resampling operation serves as a bridge that reconciles resolution differences between reference and current frames.
2Measurement precision
If motion vectors are applied without resolution alignment, then decoding complexity remains low, but motion information accuracy deteriorates
Solution Approach 1:
The patent performs preliminary resampling of reference pictures before motion compensation operations. By pre-aligning the resolution of reference pictures to match current pictures, the system ensures that subsequent motion vector applications are accurate without requiring complex real-time adjustments during decoding. This preliminary resolution alignment action eliminates the need for complex runtime resolution matching.
Solution Approach 2:
The patent modifies the resolution parameters of reference pictures through resampling to match the spatial dimensions of current pictures. This parameter change ensures that motion vectors derived from one resolution can be accurately applied to pictures at a different resolution. The resampling process adjusts the spatial parameters of reference pictures, thereby improving motion information accuracy while keeping the decoding process straightforward.
3Ease of operation
If resolution differences between reference and current pictures are not handled, then processing speed remains high, but user experience deteriorates
Solution Approach 1:
The patent changes the resolution parameters of reference pictures through resampling to match current picture dimensions. This parameter adjustment ensures consistent visual quality across frames with different resolutions, thereby improving user experience. The resampling operation maintains frame rate and processing efficiency while eliminating visual artifacts that would otherwise degrade user perception of video quality.
Data Source
AI summary
An example method of video processing includes performing a conversion between a video including a video picture including one or more video units and a bitstream representation of the video. The bitstream representation conforms to a format rule that specifies that a first syntax element is included in a header of the video picture indicating allowed prediction types of at least some of the one or more video units in the video picture.