Abstract:The paper proposes a novel error resilience protection algorithm for speech codec to enhance the reconstructed voice quality under high error rate narrowband wireless channels.At the encoder side,the proposed algorithm utilizes the resid- ual redundancy inherent in the coded voice for performing BCH and parity check to protect the parameters which are im- portant to the reconstructed speech quality.At the decoder side,the unvoiced and voiced parameters of speech codec are recovered by the modified maximum posterior transition probabilities algorithm based on the long term statistical property, and then the line spectrum pair (LSP) parameters are recovered in the voiced frame using joint source and channel char- acteristics.BCH decoding and error replacement are utilized in the unvoiced frame.The proposed algorithm requires no extra bandwidth and algorithm delay while greatly enhancing the error resilience ability of speech codec and the recon- structed voice quality under severe channel conditions.The simulations showed that the MOS (mean opinion score) was improved by 12.33% and the spectrum distortion was reduced by 25%～36.1%.

Bandpass-voicing Mode-Based Error Resilience Scheme for Vocoder Gains

Channel Error Resilience Algorithm for Vocoder LSP Parameters

Research on Bandwidth Mismatch Compensation in Speech Recognition

Error Correction Scheme for Vocoder Pitch

Consecutive Packet Loss Concealment Algorithm Based on SELP Vocoder

Channel error resilience protection algorithm for speech codec

Error Correction Scheme for SELP Vocoder Parameters

Realtime robust speech communication based on iterative joint source-channel decoding and demodulation algorithm for MELP vocoder

Interactive H.264/SVC Error Control Methods Design and Implementation

Improving G.728'S Hybrid Window And Excitation Gain

Improvement of SELP Speech Coding Model for Anti-channel-error

An Error Resilience Scheme for Very-Low-Bit-Rate Speech Codec

Studying Performance For The Gain Filters Of 8kbit/S Ld-Acelp

Improvement of Voiced-Unvoiced Classification in Vocoders

Pitch-scaled Analysis Based Residual Reconstruction for Speech Analysis and Synthesis

Low Pass Filtering and Bandwidth Extension for Robust Anti-spoofing Countermeasure Against Codec Variabilities

How to Optimize the Gain Filter of LD-CELP

High Efficient Quantization of the Energy Parameter in 0.6 Kb/s Vocoders

Joint Source-Channel Coding and Modulation Scheme for Low-Bit-rate Speech

The Statistic-Based Prediction of Excitation Spectral Parameters in Low Bit-rate Vocoder

Optimizing Gain Codebook of LD-CELP.