Deductive Refinement of Species Labelling in Weakly Labelled Birdsong Recordings
This work addresses the need for more accurate species labelling in bird sound archives and monitoring, though it appears incremental as it builds on existing classification approaches.
The paper tackles the problem of precise bird vocalization classification in weakly labelled recordings by introducing a two-step method that first detects vocalizations and then classifies them, achieving up to 89% correct classification on a synthetic dataset with vocalizations larger than 1000 pixels.
Many approaches have been used in bird species classification from their sound in order to provide labels for the whole of a recording. However, a more precise classification of each bird vocalization would be of great importance to the use and management of sound archives and bird monitoring. In this work, we introduce a technique that using a two step process can first automatically detect all bird vocalizations and then, with the use of 'weakly' labelled recordings, classify them. Evaluations of our proposed method show that it achieves a correct classification of 61% when used in a synthetic dataset, and up to 89% when the synthetic dataset only consists of vocalizations larger than 1000 pixels.