MULTIPLE FUNDAMENTAL FREQUENCY ESTIMATION BASED ON SPECTRAL PATTERN LOUDNESS AND SMOOTHNESS

Two multiple fundamental frequency estimation systems are presented in this work. In the first one (PI1, PI2), the best fundamental frequency candidates combination is found in a frame-by-frame analysis by applying a set of rules, taking into account the spectral smoothness measure described in this work. The second system (PI3) was used to extract symbolic features for audio genre classification in a fast way, so the evaluation of this system can reveal the potential of another similar approaches to support these kind of tasks.