Mixed Excitation Codebook for Natural Low-Bitrate Speech Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Low bit rate speech coding technologies face challenges in efficiently encoding GENERIC speech signals, often resulting in spiky sounds when using pulse-like codebooks and noisy sounds when using noise-like codebooks, due to the inability to effectively combine periodic and noise components.
Innovation Solution
The implementation of a pulse-noise mixed codebook structure that combines pulse-like and noise-like codebook entries, allowing for enhanced excitation vector generation through correlation and energy-based criterion functions, which improves perceptual quality by adapting to both periodic and noisy components in speech signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a pulse-like codebook is used for excitation, then periodic components are well represented, but the speech signal produces spiky sounds
Solution Approach 1:
The patent combines pulse-like codebook entries and noise-like codebook entries to form a mixed codebook structure. This merging allows the system to simultaneously represent periodic components (from pulse entries) and smooth out spiky artifacts (through noise entries), resolving the contradiction between reliable periodic representation and acceptable sound quality.
Solution Approach 2:
The excitation signal is constructed as a composite of pulse-like and noise-like components from different codebooks. This composite approach allows the system to leverage the strengths of both codebook types while mitigating their individual weaknesses, producing speech with both periodic accuracy and natural sound quality.
2Object-affected harmful factors
If a noise-like codebook is used for excitation, then smooth sound quality is achieved, but periodic components are poorly represented
Solution Approach 1:
The mixed codebook structure merges noise-like entries (providing smooth sound quality) with pulse-like entries (providing periodic representation). This combination allows the system to achieve both smooth sound quality and accurate periodic component representation simultaneously.
Solution Approach 2:
The system applies different codebook types to different local characteristics of the speech signal. Pulse-like codebook entries are used where periodic components dominate, while noise-like entries are used where smooth sound quality is prioritized, creating a locally optimized excitation signal.
3Object-affected harmful factors
If a mixed codebook structure is implemented, then perceptual quality is improved, but computational complexity increases
Solution Approach 1:
The codebook search process is segmented into multiple stages: first identifying candidate entries from pulse-like codebook, then finding complementary entries from noise-like codebook. This segmentation reduces the overall search complexity compared to exhaustively searching all possible mixed codebook combinations at once.
Solution Approach 2:
The system performs preliminary selection of candidate codebook entries based on correlation criteria before final combination. This preliminary action reduces the search space for the mixed codebook, making the overall process computationally feasible while still achieving improved perceptual quality.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
In accordance with an embodiment, a method of encoding an audio/speech signal includes determining a mixed codebook vector based on an incoming audio/speech signal, where the mixed codebook vector includes a sum of a first codebook entry from a first codebook and a second codebook entry from a second codebook. The method further includes generating an encoded audio signal based on the determined mixed codebook vector, and transmitting a coded excitation index of the determined mixed codebook vector.