Parameter Decoding Apparatus Frame Erasure Concealment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional parameter quantization methods using MA-type predictors face challenges in maintaining subjective quality during predictive quantization, particularly when frame erasure occurs, leading to excessive attenuation and degradation in sound quality due to incorrect decoding results.
Innovation Solution
A parameter decoding apparatus and method that employs a prediction residue decoding section to find a quantized prediction residue from a weighted linear sum of past and future frame parameters, enabling effective concealment processing by selecting appropriate weighting coefficients for quantized parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If predictive quantization using MA-type predictor is applied, then quantization performance is improved, but error robustness deteriorates when frame erasure occurs
Solution Approach 1:
The patent applies preliminary action by preparing multiple candidate parameters and concealment strategies in advance before frame erasure occurs. The system pre-calculates multiple possible parameter values using different prediction methods and selects the most appropriate one when erasure is detected, rather than relying on a single pre-computed parameter that may be incorrect.
Solution Approach 2:
The patent changes parameters by dynamically adjusting the prediction method and weighting coefficients based on the erasure state. When frame erasure is detected, the system switches from using the current frame's quantized prediction residue to using parameters from future frames with adjusted weighting, thereby adapting the parameter set to the erroneous condition.
2Device complexity
If conventional concealment processing is used, then decoding complexity is reduced, but subjective quality deteriorates due to excessive attenuation
Solution Approach 1:
The patent applies dynamics by making the concealment processing adaptive rather than static. The weighting coefficient applied to future frame parameters is dynamically adjusted based on the specific erasure condition and prediction context, allowing the system to optimize quality for each erasure event rather than applying fixed attenuation.
Solution Approach 2:
The patent uses feedback by utilizing information from future frames (frames after the erased frame) to correct the parameter estimation. The system feeds forward parameter information from successfully received future frames to compensate for the erased frame, creating a feedback loop that improves concealment quality without significantly increasing complexity.
3Measurement precision
If parameter interpolation from future frames is attempted, then concealment accuracy is improved, but decoding reliability worsens due to predictor errors from preceding frames
Solution Approach 1:
The patent extracts only the necessary information from future frames for concealment processing, rather than using the entire future frame parameter set. By selectively extracting and using only the specific parameter components needed for interpolation, the system reduces the propagation of errors from the predictor while maintaining concealment accuracy.
Solution Approach 2:
The patent introduces an intermediary processing step that mediates between the erased frame parameter and the future frame parameter. Instead of directly using future frame parameters, the system uses quantized prediction residues and weighted linear combinations as intermediaries to smoothly transition between frames, reducing error propagation.
Data Source
AI summary
A parameter decoding apparatus includes a prediction residue decoder that finds a quantized prediction residue based on encoded information included in a current frame subject to decoding and a moving-average predictor produces a predicted parameter by multiplying a predictive coefficient with a past quantized prediction residue. An adder decodes a parameter by adding the quantized prediction residue and the predicted parameter, wherein the prediction residue decoder, when the current frame is erased, finds a current-frame quantized prediction residue from a weighted linear sum of a parameter decoded in the past and a future-frame quantized prediction residue.