ABSTRACT
Although stethoscopes have not changed much since 1819 (Figure 4.1),62 digital technologies have evolved exponentially. In addition to sensing audio signals, we are able to record, filter, and transmit audio data. All of these functions are available in mobile phones today. Therefore, sound recognition is a “low-hanging fruit” for ambient diagnostics. This chapter covers the following:
• Modern microphones • Digital audio data formats • Sound feature representation • Sound pattern recognition
To show practical applications of the algorithms, we also study the cases of white noise measurement at the workspace and asthma sound recognition.
