Codebook Training Latency Reduction via Parallel Asset Grouping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cloud-based digital media systems face significant latency issues during codebook training, which delays the encoding and storage of digital assets, as conventional methods rely on training codebooks from a single digital asset, leading to increased processing demands and strain on cloud-based systems.
Innovation Solution
The technique involves grouping digital assets with common characteristics and processing them in parallel to train a single codebook, reducing the time required for codebook generation by leveraging the collective training data from multiple assets, thereby mitigating latency and enhancing encoding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If codebook training is performed using a single digital asset, then the training process is simple to implement, but the codebook generation latency is high and processing efficiency is low
Solution Approach 1:
The patent merges multiple digital assets with common characteristics into a single codebook training process. Instead of training separate codebooks for each asset, the system combines training data from multiple assets to generate a shared codebook, thereby reducing generation latency while maintaining compression quality.
Solution Approach 2:
The patent creates a universal codebook that can be applied to multiple digital assets sharing common characteristics. This single codebook serves multiple functions across different assets, eliminating the need for separate training processes and reducing overall system complexity.
2Loss of time
If multiple digital assets are processed in parallel to train a single codebook, then codebook generation latency is reduced, but processing resource demands increase
Solution Approach 1:
The patent segments the codebook training process by identifying and grouping digital assets with common characteristics. This segmentation allows parallel processing of grouped assets while avoiding the computational overhead of processing all assets individually, thus reducing both time and resource consumption.
Solution Approach 2:
The patent applies partial action by selecting only the essential common characteristics from multiple assets for codebook training, rather than processing all aspects of each asset. This approach reduces resource demands while still achieving significant latency reduction through parallel processing.
3Productivity
If digital assets with common characteristics are grouped and processed in parallel, then encoding efficiency is improved, but the complexity of identifying and grouping assets increases
Solution Approach 1:
The patent performs preliminary analysis to identify common characteristics among digital assets before the codebook training process. By pre-grouping assets based on their characteristics, the system simplifies the subsequent parallel processing stage and reduces the overall complexity of characteristic analysis.
Data Source
AI summary
Embodiments of the present disclosure provide systems, methods, and computer storage media for mitigating delays typically experienced when training codebooks during the encoding process. Instead of training a codebook based on a single digital asset, multiple digital assets determined to have asset characteristics in common can be grouped together to form a group of digital assets, from which a single codebook can be trained. The group of digital assets together form a codebook training set, such that each digital asset therein can be analyzed, in parallel, to expeditiously train a single codebook. A codebook trained in this manner can be employed to encode other digital assets sharing the asset characteristics as those in the codebook training set.