LPC Filter Quantization Using Multi-Reference Variable Bit Rate Coding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current speech and audio coding techniques face challenges in efficiently representing LPC filters with a variable bit rate while maintaining subjective quality, particularly in applications requiring low bit rates and mixed content encoding.

Innovation Solution

The proposed solution involves a multi-reference differential quantization method for LPC filters, where the LPC filter is differentially quantized relative to various possible references, including past or future quantized filters, or results of extrapolation/interpolation operations, with a selector choosing the reference that minimizes distortion or bit rate, and a switched absolute/differential quantization scheme to adapt to changing signal conditions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If fixed bit rate quantization is used for LPC filters, then implementation is simple, but bit rate efficiency is poor

Engineering Contradiction:
Improvebit rateVSAvoidquantization scheme complexity
Core Design Contradiction:
Quantity of substanceVSDevice complexity

Solution Approach 1:

The patent implements variable bit rate quantization where the number of bits allocated to LPC filter quantization varies dynamically based on signal characteristics and coding mode. The system switches between different quantization schemes (full quantization, differential quantization, selective quantization) depending on whether the frame contains speech or music, thereby optimizing bit rate efficiency while adapting to changing signal conditions.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes the quantization parameters (number of bits, quantization scheme type) based on the coding mode and signal properties. In music frames, LPC quantization is omitted entirely; in speech frames, variable bit allocation is applied based on the selected speech coding mode, thus adapting the parameter set to minimize overall bit rate while maintaining quality.

Inventive Principle:
Principle #35Parameter changes

2Reliability

If LPC filter is transmitted every frame, then quality is maintained, but bit rate increases

Engineering Contradiction:
Improvereconstruction qualityVSAvoidbit rate
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

The patent extracts and transmits LPC filter parameters only when necessary - specifically only in speech-coded frames and only when the selected speech mode requires it. In music frames and in frames coded with other speech modes, LPC transmission is omitted entirely, thus reducing overall bit rate while maintaining quality where it matters.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

Instead of transmitting LPC filters for every frame (excessive action), the patent applies partial quantization/transmission only to the necessary subset of frames (speech frames with appropriate coding modes). This partial action approach reduces bit rate consumption while maintaining sufficient reconstruction quality for speech segments.

Inventive Principle:
Principle #16Partial or excessive action

3Quantity of substance

If variable bit rate quantization is implemented, then bit rate efficiency improves, but complexity of quantization scheme increases

Engineering Contradiction:
Improvebit rateVSAvoidquantization scheme
Core Design Contradiction:
Quantity of substanceVSDevice complexity

Solution Approach 1:

The patent segments the quantization process into distinct schemes based on coding mode: full quantization for certain speech modes, differential quantization for other speech modes, and no quantization for music frames. This segmentation allows the system to apply the appropriate level of complexity only when needed, optimizing the trade-off between bit rate efficiency and implementation complexity.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces a mode selection mechanism that acts as an intermediary, deciding which quantization scheme to apply based on the coding mode and signal characteristics. This intermediary layer simplifies the overall system by automatically selecting the appropriate quantization approach, avoiding the need for complex real-time optimization while still achieving variable bit rate efficiency.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP2313887B1Variable bit rate LPC filter quantizing and inverse quantizing device and method
Publication Date: 2017.09.13 VOICEAGE CORPORATION
  • EP2313887B1 patent drawingFigure 1
  • EP2313887B1 patent drawingFigure 2
  • EP2313887B1 patent drawingFigure 3

AI summary

A device and a method for quantizing a LPC filter in the form of an input vector in a quantization domain, comprises a calculator of a first-stage approximation of the input vector, a subtractor of the first-stage approximation from the input vector to produce a residual vector, a calculator of a weighting function from the first-stage approximation, a warper of the residual vector with the weighting function, and a quantizer of the weighted residual vector to supply a quantized weighted residual vector. A device and a method for inverse quantizing of a LPC filter, comprises means for receiving coded indices representative of a first-stage approximation of a vector representative of the LPC filter in a quantization domain and of a quantized weighted residual version of the vector, a calculator of an inverse weighting function from the first-stage approximation, an inverse quantizer of the quantized weighted residual version of the vector to produce a weighted residual vector, a multiplier of the weighted residual vector by the inverse weighting function to produce a residual vector, and an adder of the first-stage approximation with the residual vector to produce the vector representative of the LPC filter in the quantization domain.