Speech Decoder Phase Reconstruction Using Basis Functions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing speech codecs struggle in very low bitrate scenarios due to insufficient bandwidth or transmission quality problems, leading to suboptimal speech quality in real-time communication.
Innovation Solution
The innovations involve improved speech encoding and decoding techniques, including phase quantization during encoding and phase reconstruction during decoding. Specifically, the speech encoder represents phase values using a linear component and a weighted sum of basis functions, and omits higher-frequency phase values above a cutoff frequency, while the decoder reconstructs phase values using similar methods and synthesizes omitted higher-frequency values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If conventional speech codecs are used in very low bitrate scenarios, then bandwidth consumption is reduced, but speech quality deteriorates due to insufficient bandwidth or transmission quality problems
Solution Approach 1:
The patent changes the parameter representation method by using linear prediction coefficients to model phase values instead of transmitting raw phase values. This parameter transformation allows for more efficient encoding at low bitrates while maintaining speech quality, as the linear prediction model captures the essential characteristics of phase variations with fewer bits.
Solution Approach 2:
The patent applies different encoding strategies to different parts of the speech signal. Specifically, it uses linear prediction for phase values in certain frequency ranges while using other methods for others, allowing optimized quality-bitrate tradeoff for different spectral regions. This local differentiation maintains speech quality where most important while reducing bitrate in less critical regions.
2Reliability
If more phase values are encoded to maintain speech quality, then speech quality is improved, but bitrate increases which is problematic in low bitrate scenarios
Solution Approach 1:
The patent transforms the encoding approach by representing phase values through linear prediction coefficients rather than direct phase value encoding. This parameter change reduces the number of bits required to represent phase information while maintaining the quality needed for speech reconstruction, directly addressing the quality-bitrate tradeoff.
Solution Approach 2:
The patent extracts and encodes only the essential characteristics of phase values using linear prediction, rather than encoding all phase value details. By taking out only the most important predictive parameters, the system maintains speech quality while significantly reducing the bitrate requirement for phase information transmission.
3Quantity of substance
If linear prediction coefficients are used for phase representation, then bitrate is reduced, but encoding and decoding complexity increases
Solution Approach 1:
The patent performs linear prediction coefficient calculation and phase modeling during the encoding phase as a preliminary action. By preparing and transmitting these coefficients in advance, the decoding process is simplified, as the receiver can directly use the provided coefficients for phase reconstruction without performing complex calculations. This shifts complexity to the encoding end where it can be managed more effectively.
Data Source
AI summary
Innovations in phase quantization during speech encoding and phase reconstruction during speech decoding are described. For example, to encode a set of phase values, a speech encoder omits higher-frequency phase values and/or represents at least some of the phase values as a weighted sum of basis functions. Or, as another example, to decode a set of phase values, a speech decoder reconstructs at least some of the phase values using a weighted sum of basis functions and/or reconstructs lower-frequency phase values then uses at least some of the lower-frequency phase values to synthesize higher-frequency phase values. In many cases, the innovations improve the performance of a speech codec in low bitrate scenarios, even when encoded data is delivered over a network that suffers from insufficient bandwidth or transmission quality problems.