Motion Vector Scaling Across Reference Lists in Video Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video compression methods face inaccuracies in motion vector prediction due to limited use of neighboring motion vectors from the same reference picture list, increased complexity in selecting candidates, and inefficiencies in process complexity, particularly when considering temporal differences between reference pictures.
Innovation Solution
The method involves scaling neighboring motion vectors based on temporal distance differences between reference pictures, extending candidate groups to include vectors from different reference picture lists, and optimizing priority search conditions for accurate motion vector prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If neighboring motion vectors from the same reference picture list are used for prediction, then the prediction process is simple, but the prediction accuracy decreases
Solution Approach 1:
The patent extends the motion vector candidate selection from a single reference picture list to multiple reference picture lists (List 0 and List 1). This dimensional expansion allows the system to search for neighboring motion vectors across different reference lists, thereby improving prediction accuracy without significantly increasing process complexity.
Solution Approach 2:
The patent introduces temporal distance as a scaling parameter to adjust neighboring motion vectors. By calculating the temporal distance between the current picture and the reference picture, and scaling the motion vector accordingly, the system improves prediction accuracy for vectors from different reference pictures while maintaining computational efficiency.
2Adaptability or versatility
If neighboring motion vectors from different reference picture lists are included, then the selection range is expanded, but the search complexity increases
Solution Approach 1:
The patent applies different selection criteria to different neighboring partitions based on their availability and reliability. High-priority partitions (A0, A1, B0, B1) are evaluated first with stricter conditions, while lower-priority partitions are considered only if higher-priority ones are unavailable. This localized quality approach expands the selection range while managing search complexity through priority-based filtering.
Solution Approach 2:
The patent performs preliminary checks on reference picture list indices and temporal distances before incorporating neighboring motion vectors into the candidate group. By pre-calculating temporal distance scaling factors and verifying reference list compatibility in advance, the system expands its selection range while avoiding redundant computations during the main prediction process.
3Measurement precision
If temporal distance scaling is applied to neighboring motion vectors, then the prediction accuracy for different reference pictures improves, but the computational complexity increases
Solution Approach 1:
The patent scales neighboring motion vectors by the ratio of temporal distances (temporal distance from current picture to reference picture divided by temporal distance from reference picture to its source). This parameter transformation adjusts vectors from different reference pictures to match the current picture's temporal context, improving accuracy while using a simple multiplicative scaling operation that is computationally efficient.
Solution Approach 2:
The patent applies temporal distance scaling selectively only to neighboring motion vectors from reference pictures that are not the current picture's direct reference. Vectors from the primary reference list require no scaling, while only vectors from alternative reference lists undergo the scaling operation. This partial application reduces computational complexity while still improving accuracy where needed.
Data Source
AI summary
The present invention relates to a method and apparatus for processing a video signal, which can increase the accuracy of the motion vector prediction through motion vector scaling which takes a difference in the temporal distance between reference pictures into consideration. To this end, the present invention provides a video signal processing method and a video signal processing apparatus using the same, and the method comprises the steps of: scaling at least one neighboring partition motion vector for a motion vector prediction of the current partition; scaling the neighboring partition motion vector, which has been selected, when the reference picture of the neighboring partition motion vector is different from the reference picture of the current partition; acquiring a motion vector prediction value of the current partition using the scaled motion vector; and acquiring a motion vector of the current partition using the motion vector prediction value.