Speech Decoder Pitch Lag Adjustment for Error Frames
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing speech decoding technologies face issues with excessive periodicity and reduced decoding accuracy due to continuous frame errors, particularly in ACELP-based speech decoders, where pitch lag parameters are not effectively adjusted, leading to sharp noise and quality deterioration.
Innovation Solution
A method and apparatus that calculate and adjust the pitch lag parameter of erroneous frames by fluctuating it around the previous frame's value based on the number of continuous errors, ensuring the parameter remains within a preset range, thereby reducing accumulated errors and avoiding excessive periodicity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the pitch lag parameter of the previous frame is used repeatedly for erroneous frames, then the decoding process is simple and fast, but excessive periodicity occurs and decoding accuracy deteriorates
Solution Approach 1:
The pitch lag parameter is made dynamic by introducing random fluctuation around the previous frame's value instead of using a fixed repeated value. The formula τ(m) = τ(m-1) + Δτ where Δτ is a random value within [-δ, δ] creates a dynamic adaptation mechanism that responds to frame errors while maintaining natural speech variability, thus improving accuracy without significantly compromising decoding speed.
Solution Approach 2:
The invention changes the parameter determination strategy from static (repeating previous value) to stochastic (adding random fluctuation). By modifying how the pitch lag parameter is calculated during error frames - introducing a controlled random component - the system maintains simplicity while improving accuracy by preventing excessive periodicity.
2Stability of the object's composition
If the pitch lag parameter is increased by 1 for erroneous frames, then excessive periodicity is avoided, but accumulated error increases and decoding accuracy reduces
Solution Approach 1:
Instead of monotonically increasing the pitch lag parameter by fixed amounts, the invention introduces random fluctuation around the previous frame's value. This stochastic approach prevents systematic drift and accumulated error while still avoiding excessive periodicity, thereby maintaining both stability and accuracy.
Solution Approach 2:
The invention uses periodic random fluctuation rather than systematic increment. By adding random values within a bounded range [-δ, δ] at each error frame, the system creates a bounded periodic-like behavior that prevents both excessive periodicity and accumulated error, maintaining parameter accuracy over multiple erroneous frames.
3Measurement precision
If signal classification is performed to determine pitch lag parameters, then decoding accuracy improves, but system complexity increases
Solution Approach 1:
The invention extracts only the essential element needed for accuracy improvement - the random fluctuation component - and applies it directly to the pitch lag parameter calculation. This avoids the need for complex signal classification while still achieving better accuracy, by taking out and applying only the necessary stochastic adjustment.
Solution Approach 2:
Rather than implementing complex signal classification systems, the invention achieves accuracy improvement through a simple parameter change - adding random fluctuation to the pitch lag parameter. This lightweight approach maintains low system complexity while effectively improving decoding accuracy by preventing excessive periodicity.
Data Source
AI summary
The present disclosure relates to a decoding method and apparatus. The method includes: receiving data frames from the coder; if any erroneous frame appears, calculating a pitch lag parameter of the erroneous frame; decoding the data frames according to the calculated pitch lag parameter of the erroneous frame, and obtaining decoded data. The process of determining the pitch lag parameter includes: determining the number of continuous erroneous frames and the pitch lag parameter of the previous frame; adjusting the pitch lag parameter of the previous frame according to the number of the continuous erroneous frames and a preset adjustment policy, and calculating and determining the pitch lag parameter of a current erroneous frame, wherein the preset adjustment policy is adjusting the determined pitch lag parameter of the current erroneous frame within a preset value range according to the number of the continuous erroneous frames.