Bird Species Identification Based on Timbre and Pitch Features of Their Vocalization

Abstract

Although wild bird watching has been a popular leisure activity, it is often the case that people have no idea about what kind of bird species they see or hear. To help people learn to identify bird species, this study proposes an automatic bird sound identification system. Considering bird vocalization can be generally divided into two categories, namely call and song, the proposed system is built upon a two-stage identification framework. The first stage performs call/song classification. If an unknown sound clip is classified as a call, it is then handled by a call identifier in the second stage; otherwise, it is handled by a song identifier. Both identifiers use two acoustic features, timbre and pitch, to determine which of the bird species the sound clip belong to. In using timbre features, bird sounds are converted into Mel-Frequency Cepstral Coefficients and their first derivatives and then analyzed using Gaussian mixture models. In using pitch feature, we convert bird sounds into MIDI note sequences and then use bigram models to analyze the dynamic change information of the notes. Our experiments, conducted using a database including twenty common bird species in the Taipei urban area, show that the proposed system can achieve 90.4% accuracy.