Adaptive Motion Vector Ordering for Lower Video Coding Overhead
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards like MPEG-4 AVC consume excessive bits in transmitting motion vector predictor indices due to limited information-based ordering of motion vector candidates.
Innovation Solution
Adaptive motion vector candidate ordering is implemented using common information available at both encoder and decoder, such as selection frequency, block type, consistency, fidelity, and reference index, to assign smaller indices to frequently selected predictors, reducing overhead bits.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If motion vector predictor indices are transmitted explicitly to convey multiple candidates, then prediction accuracy is improved, but overhead bits increase significantly
Solution Approach 1:
The patent applies dynamics by making the candidate set and its ordering adaptive rather than fixed. The candidate set is dynamically constructed based on block characteristics (e.g., motion mode, prediction mode) and the ordering is dynamically adjusted based on encoder/decoder context. This allows the system to optimize between prediction accuracy and bit rate for each specific coding situation, resolving the contradiction by making the information transmission efficient without sacrificing accuracy.
Solution Approach 2:
The patent changes the parameter of candidate ordering from a fixed conventional order to an adaptive order based on multiple characteristics including block partition mode, motion mode, and prediction mode. By changing the ordering parameter dynamically based on these characteristics, the system can reduce the number of bits needed while maintaining or improving prediction accuracy, as the most relevant candidates are positioned first in the adaptive ordering.
2Device complexity
If a fixed ordering of motion vector candidates is used, then implementation complexity is reduced, but coding efficiency deteriorates
Solution Approach 1:
The patent applies local quality by tailoring the candidate ordering to specific local conditions (block characteristics) rather than using a uniform fixed ordering. Different block types, motion modes, and prediction modes trigger different ordering strategies. This localized adaptation improves coding efficiency for each specific case without requiring a completely complex global optimization system, thus balancing implementation complexity with coding efficiency.
Solution Approach 2:
The patent performs preliminary action by pre-defining multiple possible orderings based on block characteristics before the actual encoding process. The encoder determines the appropriate ordering in advance based on characteristics like partition mode and motion mode, and applies it consistently. This preliminary setup avoids the need for complex real-time optimization while still achieving high coding efficiency through pre-planned adaptive orderings.
3Measurement precision
If more motion vector candidates are considered, then prediction accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent applies partial action by considering only the necessary number of candidates based on the specific block characteristics and coding context. Instead of always considering all possible motion vector candidates, the system selects and orders a manageable subset that is sufficient for achieving good prediction accuracy in each case. This partial consideration reduces computational complexity while maintaining adequate prediction accuracy.
Solution Approach 2:
The patent makes the candidate set size and ordering dynamic based on block characteristics. For blocks with simple characteristics, a smaller candidate set with a specific ordering is used, reducing computational complexity. For blocks with more complex characteristics, a larger candidate set with optimized ordering is used, improving prediction accuracy. This dynamic adaptation allows the system to balance computational complexity and prediction accuracy according to the specific coding situation.
Data Source
AI summary
Methods and apparatus are provided for adaptive motion vector candidate ordering for video encoding and decoding. An apparatus includes a video encoder (100) for encoding a block in a picture by selecting an order of motion vector predictor candidates for the block responsive to a characteristic available at both the video encoder and a corresponding decoder. The characteristic excludes a mode in which the block is partitioned.