Entropy Decoder Context Switching for Variable Data Streams
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing entropy decoding methods are inflexible and lack adaptability when dealing with data formats that are not fully predefined, particularly in contexts where the technical nature of the data depends on artificial neural networks, such as those encoding or decoding data in multiple languages or varying alphabets, or when the association of data to entropy encoding contexts cannot be defined in advance.
Innovation Solution
A decoding method and device that utilize identifiers in the data stream to determine contexts for entropy decoding, allowing flexibility by using machine learning methods like deep learning or random forest learning, and configuring entropy decoders based on these contexts, along with an electronic device that includes an entropy decoder and a configuration module to parameterize the decoder accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional entropy decoding methods are used with predefined contexts, then decoding speed is maintained, but adaptability to varying data formats deteriorates
Solution Approach 1:
The patent implements dynamic context selection where the entropy decoder can adaptively choose different contexts based on identifiers in the data stream. Instead of using fixed predefined contexts, the system dynamically determines which context model to apply for each syntax element based on the current data characteristics, enabling adaptation to varying data formats while maintaining efficient decoding.
Solution Approach 2:
The patent changes the parameterization of the entropy decoder by introducing context identifiers that select from multiple context models. Each context model has different probability parameters adapted to specific data types or formats. By changing which context parameters are active based on stream identifiers, the system achieves format adaptability without requiring complete redesign of the decoder architecture.
2Adaptability or versatility
If multiple context models are introduced to handle different data types, then adaptability improves, but processing time increases
Solution Approach 1:
The patent prepares multiple context models in advance, each optimized for specific data types or formats. During decoding, the system quickly selects the appropriate pre-prepared context based on identifiers in the stream, avoiding the need to create or adapt contexts on-the-fly. This preliminary preparation of multiple specialized contexts enables fast switching between data types without significant processing overhead.
3Adaptability or versatility
If context identifiers are added to the data stream, then flexibility in entropy decoding improves, but data stream complexity increases
Solution Approach 1:
The patent designs the context identifier field to serve multiple functions: it identifies the appropriate context model for entropy decoding, indicates the data type being processed, and can signal formatting information. This multi-functional use of the identifier reduces the need for separate metadata fields, achieving flexibility in entropy decoding without proportionally increasing data stream complexity.
Data Source
AI summary
A decoding method implemented by a decoding device. The decoding method includes: obtaining a data stream, including a plurality of identifiers, a sequence of binary elements and an indicator indicating a start of the sequence of binary elements within the data stream; and decoding the obtained data stream into a sequence of data of respective predetermined types. The decoding includes: obtaining at least one piece of data of the sequence of data by: determining a context based on an identifier, from among the plurality of identifiers, that is associated with the respective type of the respective piece of data; identifying the start of the sequence of binary elements from the indicator; and decoding a part of the sequence of binary elements by using an entropy decoder receiving the sequence of binary elements as an input and parameterized in the determined context.