Power pooling: An adaptive pooling function for weakly labelled sound event detection
This work addresses the challenge of sound event detection in engineering applications where strong labels are scarce, offering an incremental improvement over existing pooling functions.
The paper tackles the problem of detecting sound events with weak labels by proposing an adaptive power pooling function for multiple instance learning, which improves event-based F1 scores by 11.4% and 10.2% on two datasets compared to state-of-the-art methods.
Access to large corpora with strongly labelled sound events is expensive and difficult in engineering applications. Much research turns to address the problem of how to detect both the types and the timestamps of sound events with weak labels that only specify the types. This task can be treated as a multiple instance learning (MIL) problem, and the key to it is the design of a pooling function. In this paper, we propose an adaptive power pooling function which can automatically adapt to various sound sources. On two public datasets, the proposed power pooling function outperforms the state-of-the-art linear softmax pooling on both coarsegrained and fine-grained metrics. Notably, it improves the event-based F1 score (which evaluates the detection of event onsets and offsets) by 11.4% and 10.2% relative on the two datasets. While this paper focuses on sound event detection applications, the proposed method can be applied to MIL tasks in other domains.