Mixed Codebook Excitation for Low-Bitrate Speech Quality
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Low bit rate speech coding technologies face challenges in efficiently encoding GENERIC class speech signals, as they often result in spiky sounds when using pulse-like codebooks and noisy sounds when using noise-like codebooks, due to the mixed periodic and noise components in these signals.
Innovation Solution
A pulse-noise mixed codebook structure is introduced, combining pulse-like and noise-like codebook entries to generate a mixed codebook vector, which is used to create a coded excitation for speech encoding, allowing for better waveform matching and perceptual quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If pulse-like codebook entries are used for encoding GENERIC class speech signals, then the encoding efficiency is improved, but the output quality deteriorates with spiky sounds
Solution Approach 1:
The patent combines pulse-like codebook entries and noise-like codebook entries into a unified codebook structure. The excitation signal is formed by adding a pulse component and a noise component, where the pulse component captures periodicity and the noise component smooths the spiky artifacts. This merging allows the system to maintain encoding efficiency while improving output quality by eliminating spiky sounds.
2Object-affected harmful factors
If noise-like codebook entries are used for encoding GENERIC class speech signals, then the output quality improves with smoother sound, but the encoding efficiency deteriorates with noisy sounds
Solution Approach 1:
The patent applies different codebook entry types to different portions of the excitation signal based on local signal characteristics. The pulse component is applied where periodicity is present, and the noise component is applied where smoothing is needed. This local differentiation allows the system to achieve smooth sound quality without sacrificing overall encoding efficiency, as each component is optimized for its specific function.
3Loss of substance
If traditional codebooks are used for speech coding, then the bit rate is kept low, but the perceptual quality deteriorates
Solution Approach 1:
The patent creates a composite excitation signal by combining pulse-like codebook entries and noise-like codebook entries. This composite approach allows the system to maintain low bit rate operation while achieving improved perceptual quality. The pulse component provides structural information efficiently, while the noise component adds perceptual smoothness, together creating a low bit rate solution with high perceptual quality.
Data Source
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.