An integrated solution for error concealment in DSR systems over wireless channels

Distributed Speech Recognition involves the development of techniques to conceal the degradations that the transmission channel introduces in the speech features. This work proposes a lowcomplexity high-accuracy error concealment technique compatible with the DSR ETSI standards. This is achieved by combining three different techniques: fast MMSE estimation, Viterbi decoding with soft-data and subvector-based error detection. We also propose a method to extend this Viterbi decoding to dynamic features. The experimental results show the effectiveness of our proposal.