TriA Pipeline: A Large-Scale Automatic Audio Annotation Pipeline For Audio Classification In Specific Scenarios
This work addresses the problem of limited annotated audio data for specific scenarios like domestic environments, providing a scalable automatic annotation method.
The authors propose the TriA Pipeline, an automatic audio annotation pipeline that converts audio from various scenarios into high-quality training data. Using this pipeline, they constructed a dataset of over 2130 hours covering 431 audio classes, and showed that augmenting manually annotated data with a prior-knowledge-guided subset yields average relative gains of 3.97% in accuracy and 3.35% in Macro-F1 on three domestic audio classification tasks.
There are some datasets of varying scales for audio classification (AC) applied to different tasks. However, annotated data is limited for most scenarios, such as domestic environments. To address this challenge, we propose an $\textbf{A}$utomatic $\textbf{A}$udio $\textbf{A}$nnotation Pipeline--TriA Pipeline, which can efficiently convert audio from various scenarios into high-quality training data with audio event annotations. A TriA dataset was constructed with the TriA Pipeline, over 2130 hours of audio covering 431 audio classes. Furthermore, we partitioned a prior-knowledge-guided subset (TriA$_{\mathrm{GK}}$) from TriA and conduct comparative experiments on three domestic AC tasks. Comparing the result on manually annotated data only and that on manually annotated data combines TriA$_{\mathrm{GK}}$, TriA$_{\mathrm{GK}}$ could achieve average relative gains of 3.97% in accuracy and 3.35% in Macro-F1, validating the effectiveness of TriA$_{\mathrm{GK}}$ and the TriA Pipeline.