Transition Mode Codebook for Speech Signal Error Propagation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

CELP-based speech coders face issues with error propagation and reduced coding efficiency due to the use of adaptive codebooks, particularly during frame erasures, especially in transition frames between unvoiced and voiced segments or between different voiced segments, leading to significant distortion in synthesized speech.

Innovation Solution

The implementation of a transition mode encoding technique that replaces the adaptive codebook with a glottal-shape codebook, which is independent of past excitation, to generate codevectors for transition frames and subsequent frames, ensuring rapid convergence to clean-channel synthesis and improved coding efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If an adaptive codebook is used in CELP coding to achieve good speech quality, then coding efficiency is improved, but error propagation occurs during frame erasures and transition frames

Engineering Contradiction:
Improvecoding efficiencyVSAvoiderror propagation
Core Design Contradiction:
Manufacturing precisionVSReliability

Solution Approach 1:

The patent segments the excitation signal representation into two independent parts: a fixed codebook providing periodic excitation and an adaptive codebook providing innovation. This segmentation allows the periodic part to be reliably reconstructed even when the adaptive codebook fails during erasures, thus reducing error propagation while maintaining coding efficiency.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the parameter representation by using a fixed codebook index for the periodic excitation instead of relying solely on adaptive codebook parameters. This parameter change ensures that the decoder can reconstruct the excitation signal correctly even when transmission errors occur, as the fixed codebook provides a stable reference.

Inventive Principle:
Principle #35Parameter changes

2Manufacturing precision

If an adaptive codebook is used to model voiced speech segments, then speech quality is improved, but convergence is slow after frame erasures

Engineering Contradiction:
Improvespeech qualityVSAvoidconvergence time
Core Design Contradiction:
Manufacturing precisionVSLoss of time

Solution Approach 1:

The patent performs preliminary action by pre-defining a fixed codebook of excitation signals that can be directly used for reconstruction. When frame erasures occur, the decoder can immediately use this pre-prepared fixed codebook without needing to wait for the adaptive codebook to converge, thus reducing convergence time while maintaining speech quality.

Inventive Principle:
Principle #10Preliminary action

3Ease of operation

If noise-like excitation from previous unvoiced frames is used to update the past excitation buffer, then the buffer is updated, but the periodic part of excitation is missing in transition frames

Engineering Contradiction:
Improvebuffer updatingVSAvoidperiodic excitation accuracy
Core Design Contradiction:
Ease of operationVSManufacturing precision

Solution Approach 1:

The patent introduces a fixed codebook as an intermediary between the noise-like excitation and the periodic excitation requirement. The fixed codebook provides the missing periodic structure in transition frames, acting as a mediator that bridges the gap between the updated buffer content and the required periodic excitation characteristics.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP2102619B1Method and device for coding transition frames in speech signals
Publication Date: 2017.03.22 VOICEAGE CORPORATION
  • EP2102619B1 patent drawingFigure 1a~1b
  • EP2102619B1 patent drawingFigure 2~3
  • EP2102619B1 patent drawingFigure 4~5c

AI summary

There is provided a transition mode device and method for use in a predictive-type sound signal codec for producing a transition mode excitation replacing an adaptive codebook excitation in a transition frame and/or a frame following the transition in the sound signal, comprising an input for receiving a codebook index and a transition mode codebook for generating a set of codevectors independent from past excitation. The transition mode codebook is responsive to the index for generating, in the transition frame and/or frame following the transition, one of the codevectors of the set corresponding to the transition mode excitation. There is also provided an encoding device and method and a decoding device and method using the above described transition mode device and method.