Beyond Equal-Length Snippets: How Long is Sufficient to Recognize an Audio Scene?
This work addresses the problem of optimizing audio scene classification efficiency for applications like surveillance or smart devices, but it is incremental as it builds on existing models and datasets.
The study investigated how long an audio clip needs to be for reliable scene recognition, finding that some scenes can be identified in a few seconds while others require much longer durations, and that model fusion is most beneficial for short clips.
Due to the variability in characteristics of audio scenes, some scenes can naturally be recognized earlier than others. In this work, rather than using equal-length snippets for all scene categories, as is common in the literature, we study to which temporal extent an audio scene can be reliably recognized given state-of-the-art models. Moreover, as model fusion with deep network ensemble is prevalent in audio scene classification, we further study whether, and if so, when model fusion is necessary for this task. To achieve these goals, we employ two single-network systems relying on a convolutional neural network and a recurrent neural network for classification as well as early fusion and late fusion of these networks. Experimental results on the LITIS-Rouen dataset show that some scenes can be reliably recognized with a few seconds while other scenes require significantly longer durations. In addition, model fusion is shown to be the most beneficial when the signal length is short.