TY - JOUR
T1 - Few-shot continual learning for audio classification
AU - Wang, Yu
AU - Bryan, Nicholas J.
AU - Cartwright, Mark
AU - Bello, Juan Pablo
AU - Salamon, Justin
N1 - Funding Information:
∗This work was performed while interning at Adobe Research. This work is partially supported by National Science Foundation award 1544753.
Publisher Copyright:
© 2021 IEEE
PY - 2021
Y1 - 2021
N2 - Supervised learning for audio classification typically imposes a fixed class vocabulary, which can be limiting for real-world applications where the target class vocabulary is not known a priori or changes dynamically. In this work, we introduce a few-shot continual learning framework for audio classification, where we can continuously expand a trained base classifier to recognize novel classes based on only few labeled data at inference time. This enables fast and interactive model updates by end-users with minimal human effort. To do so, we leverage the dynamic few-shot learning technique and adapt it to a challenging multi-label audio classification scenario. We incorporate a recent state-of-the-art audio feature extraction model as a backbone and perform a comparative analysis of our approach on two popular audio datasets (ESC-50 and AudioSet). We conduct an in-depth evaluation to illustrate the complexities of the problem and show that, while there is still room for improvement, our method outperforms three baselines on novel class detection while maintaining its performance on base classes.
AB - Supervised learning for audio classification typically imposes a fixed class vocabulary, which can be limiting for real-world applications where the target class vocabulary is not known a priori or changes dynamically. In this work, we introduce a few-shot continual learning framework for audio classification, where we can continuously expand a trained base classifier to recognize novel classes based on only few labeled data at inference time. This enables fast and interactive model updates by end-users with minimal human effort. To do so, we leverage the dynamic few-shot learning technique and adapt it to a challenging multi-label audio classification scenario. We incorporate a recent state-of-the-art audio feature extraction model as a backbone and perform a comparative analysis of our approach on two popular audio datasets (ESC-50 and AudioSet). We conduct an in-depth evaluation to illustrate the complexities of the problem and show that, while there is still room for improvement, our method outperforms three baselines on novel class detection while maintaining its performance on base classes.
KW - Audio classification
KW - Continual learning
KW - Few-shot learning
KW - Supervised learning
UR - http://www.scopus.com/inward/record.url?scp=85111216122&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85111216122&partnerID=8YFLogxK
U2 - 10.1109/ICASSP39728.2021.9413584
DO - 10.1109/ICASSP39728.2021.9413584
M3 - Conference article
AN - SCOPUS:85111216122
SN - 1520-6149
VL - 2021-June
SP - 321
EP - 325
JO - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
JF - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
T2 - 2021 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2021
Y2 - 6 June 2021 through 11 June 2021
ER -