Automatic Organisation, Segmentation, and Filtering of User-Generated Audio Content
This work addresses the challenge of managing noisy user-generated audio data for applications like event analysis, but it appears incremental as it builds on existing audio fingerprinting techniques.
The paper tackles the problem of organizing and segmenting large datasets of user-generated audio content by grouping files containing common audio excerpts and analyzing their temporal and quality correlations, validated on manually crawled YouTube concert recordings.
Using solely the information retrieved by audio fingerprinting techniques, we propose methods to treat a possibly large dataset of user-generated audio content, that (1) enable the grouping of several audio files that contain a common audio excerpt (i.e., are relative to the same event), and (2) give information about how those files are correlated in terms of time and quality inside each event. Furthermore, we use supervised learning to detect incorrect matches that may arise from the audio fingerprinting algorithm itself, whilst ensuring our model learns with previous predictions. All the presented methods were further validated by user-generated recordings of several different concerts manually crawled from YouTube.