Noise Suppression Method by Jointly Using Bone- and Air- Conducted Speech Signals

Speech recognition systems have been applied in various fields due to recent development of digital signal processing technique. For example, these systems are applied to inspection and maintenance operations in industrial factories and to recording and reporting routines at construction sites, etc. where hand-writing is difficult. For speech recognition in such actual circumstances, some countermeasure methods for surrounding noises are indispensable. Several noise suppression methods for the speech signal have been proposed up to now. In this study, a signal processing method to remove the noise for actual speech signal is proposed by jointly using the measured data of bone- and air- conducted speech signals. More specifically, by introducing Bayes' theorem based on an orthogonal expansion expression of probability distribution, a new type of noise removal method is proposed. Though the bone-conducted speech signal is a kind of solid propagation sound with less effect by the surrounding noise, the high frequency component of the signal reduces through the propagation process. On the other hand, air-conducted speech signal contains all frequency components though the signal is strongly affected by the surrounding noise. Therefore, by using jointly both of bone- and air- conducted speech signals, more accurate estimation of the speech signal can be expected with recovering the high-frequency components of the speech signal even in a very noisy circumstance. The effectiveness of the proposed method is confirmed by applying it to bone- and air- conducted speech measured in real environment under existence of surrounding noise.