Event-related data conditioning for acoustic event classification

TL;DR

This paper introduces event-related data conditioning (EDC), a novel spectrogram-based method that adaptively emphasizes local acoustic features to improve acoustic event classification, outperforming existing augmentation and attention techniques.

Contribution

The paper proposes EDC, a new data conditioning approach that adaptively selects attention ranges based on acoustic features to enhance event boundaries and classification accuracy.

Findings

EDC outperforms existing data augmentation and attention methods.

EDC effectively enhances event-background boundaries.

EDC improves AEC performance in various modes.

Abstract

Models based on diverse attention mechanisms have recently shined in tasks related to acoustic event classification (AEC). Among them, self-attention is often used in audio-only tasks to help the model recognize different acoustic events. Self-attention relies on the similarity between time frames, and uses global information from the whole segment to highlight specific features within a frame. In real life, information related to acoustic events will attenuate over time, which means the information within some frames around the event deserves more attention than distant time global information that may be unrelated to the event. This paper shows that self-attention may over-enhance certain segments of audio representations, and smooth out the boundaries between events representations and background noises. Hence, this paper proposes an event-related data conditioning (EDC) for AEC. EDC directly works on spectrograms. The idea of EDC is to adaptively select the frame-related attention range based on acoustic features, and gather the event-related local information to represent the frame. Experiments show that: 1) compared with spectrogram-based data augmentation methods and trainable feature weighting and self-attention, EDC outperforms them in both the original-size mode and the augmented mode; 2) EDC effectively gathers event-related local information and enhances boundaries between events and backgrounds, improving the performance of AEC.