14th International Congress of Phonetic Sciences (ICPhS-14)
San Francisco, CA, USA
An Italian speaker-independent continuous-speech digit recognizer is described. The CSLU Toolkit was used to develop and implement the system. In the first set of experiments, the SPK-IRST corpus, a collection of digit sentences recorded in a clean environment, was used both for training and testing the system. In the second set, a bandfiltered version (between 300 Hz and 3400 Hz) of the SPKIRST corpus was considered for training, while the telephone PANDA-CSELT corpus was used for testing the system. A hybrid HMM/NN architecture was applied; in this architecture, a three-layer neural network is used as a state emission probability estimator and the conventional forwardbackward algorithm is applied for estimating continuous targets for the NN training patterns. The final network, trained to estimate the probability of 116 context-dependent phonetic categories at every 10-msec frame, was not trained on binary target values, but on the probabilities of each phonetic category belonging to each frame. Training and testing will be described in detail and recognition results will be illustrated.
Bibliographic reference. Cosi, Piero / Hosom, John-Paul (1999): "HMM/neural network-based system for Italian continuous digit recognition", In ICPhS-14, 1669-1672.