A meta-learning approach for user-defined spoken term classification with varying classes and examples

Chen，Yangbin1; Ko，Tom2; Wang，Jianping3

10.21437/Interspeech.2021-147

2021

2308-457X

1990-9772

Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

6

4071-4075

Recently we formulated a user-defined spoken term classification task as a few-shot learning task and tackled the task using Model-Agnostic Meta-Learning (MAML) algorithm. Our results show that the meta-learning approach performs much better than conventional supervised learning and transfer learning in the task, especially with limited training data. In this paper, we extend our work by addressing a more practical problem in the user-defined scenario where users can define any number of spoken terms and provide any number of enrollment audio examples for each spoken term. From the perspective of fewshot learning, this is an N-way, K-shot problem with varying N and K. In our work, we relax the values of N and K of each meta-task during training instead of assigning fixed values to them, which differs from what most meta-learning algorithms do. We adopt a metric-based meta-learning algorithm named Prototypical Networks (ProtoNet) as it avoids exhaustive fine-tuning when N varies. Furthermore, we use the Max-Mahalanobis Center (MMC) loss as an effective regularizer to address the problem of ProtoNet under the condition of varying K. Experiments on the Google Speech Commands dataset demonstrate that our proposed method outperforms the conventional N-way, K-shot setting in most testing tasks.

Few-shot learning Meta-learning Spoken term classification

通讯

英语

EI

20214711194616

Learning systems ; Speech communication ; Statistical tests

Machine Learning:723.4.2 ; Speech:751.5 ; Mathematical Statistics:922.2

2-s2.0-85119300018

Scopus

1.Department of Information Engineering,The Chinese University of Hong Kong,China
2.Department of Computer Science and Engineering,Southern University of Science and Technology,China
3.Department of Computer Science,City University of Hong Kong,China

Chen，Yangbin,Ko，Tom,Wang，Jianping. A meta-learning approach for user-defined spoken term classification with varying classes and examples[C],2021:4071-4075.