Representation Learning for the Automatic Indexing of Sound Effects Libraries
This work addresses the time-consuming and complex task of managing commercial sound effects libraries, offering an incremental improvement through dataset-independent embeddings.
The paper tackles the problem of labeling and searching sound effects libraries by developing a representation learning approach that creates generalized, taxonomy-agnostic embeddings, outperforming established methods like OpenL3 in handling data issues such as class imbalance and inconsistent labels.
Labeling and maintaining a commercial sound effects library is a time-consuming task exacerbated by databases that continually grow in size and undergo taxonomy updates. Moreover, sound search and taxonomy creation are complicated by non-uniform metadata, an unrelenting problem even with the introduction of a new industry standard, the Universal Category System. To address these problems and overcome dataset-dependent limitations that inhibit the successful training of deep learning models, we pursue representation learning to train generalized embeddings that can be used for a wide variety of sound effects libraries and are a taxonomy-agnostic representation of sound. We show that a task-specific but dataset-independent representation can successfully address data issues such as class imbalance, inconsistent class labels, and insufficient dataset size, outperforming established representations such as OpenL3. Detailed experimental results show the impact of metric learning approaches and different cross-dataset training methods on representational effectiveness.