Robust formant tracking for continuous speech with speaker variability

Several algorithms have been developed for extracting formant information from speech signals, however most of these algorithms are either not robust in real-life noise environments or are not suitable for real-time implementation. The algorithm proposed here attempts formant extraction from continuous speech by using a set of time-varying adaptive filters to track and estimate individual formant frequencies. The algorithm incorporates an adaptive voicing detector and a gender detector for robust formant extraction from continuous speech for both male and female speakers in the presence of background noise. Initial testing of the algorithm using TIMIT database and synthesized speech sentences shows promising results in the presence of AWGN and a competing background speaker at various SNRs.