Leveraging Hierarchical Structures for Few-Shot Musical Instrument Recognition
| Main Authors: | Hugo F Flores Garcia, Aldo Aguilar, Ethan Manilow, Bryan Pardo |
|---|---|
| Format: | Proceeding Journal |
| Terbitan: |
ISMIR
, 2021
|
| Online Access: |
https://zenodo.org/record/5624615 |
Daftar Isi:
- Deep learning work on musical instrument recognition has generally focused on instrument classes for which we have abundant data. In this work, we exploit hierarchical relationships between instruments in a few-shot learning setup to enable classification of a wider set of musical instruments, given a few examples at inference. We apply a hierarchical loss function to the training of prototypical networks, combined with a method to aggregate prototypes hierarchically, mirroring the structure of a predefined musical instrument hierarchy. These extensions require no changes to the network architecture and new levels can be easily added or removed. Compared to a non-hierarchical few-shot baseline, our method leads to a significant increase in classification accuracy and significant decrease in mistake severity on instrument classes unseen in training.