Meta-Learning for Short Utterance Speaker Recognition with Imbalance Length Pairs

Abstract

In practical settings, a speaker recognition system needs to identify a speaker given a short utterance, while the enrollment utterance may be relatively long. However, existing speaker recognition models perform poorly with such short utterances. To solve this problem, we introduce a meta-learning framework for imbalance length pairs. Specifically, we use a Prototypical Networks and train it with a support set of long utterances and a query set of short utterances of varying lengths. Further, since optimizing only for the classes in the given episode may be insufficient for learning discriminative embeddings for unseen classes, we additionally enforce the model to classify both the support and the query set against the entire set of classes in the training set. By combining these two learning schemes, our model outperforms existing state-of-the-art speaker verification models learned with a standard supervised learning framework on short utterance (1-2 seconds) on the VoxCeleb datasets. We also validate our proposed model for unseen speaker identification, on which it also achieves significant performance gains over the existing approaches. The codes are available at https://github.com/seongmin-kye/meta-SR.

Accepted to Interspeech 2020. The codes are available at https://github.com/seongmin-kye/meta-SR

References in corpus (11)

• Prototypical Networks for Few-shot Learning
• VoxCeleb2: Deep Speaker Recognition
• Attentive Statistics Pooling for Deep Speaker Embedding
• Learning to Propagate Labels: Transductive Propagation Network for Few-shot Learning
• Deep Speaker: an End-to-End Neural Speaker Embedding System
• Improving Multi-Scale Aggregation Using Feature Pyramid Module for Robust Speaker Verification of Variable-Duration Utterances
• Few Shot Speaker Recognition using Deep Neural Networks
• Spatial Pyramid Encoding with Convex Length Normalization for Text-Independent Speaker Verification
• Meta-Learned Confidence for Few-shot Learning
• Intra-class variation reduction of speaker representation in disentanglement framework
• MIRNet: Learning multiple identities representations in overlapped speech

Cited by in corpus (8)

• Improving Multi-Scale Aggregation Using Feature Pyramid Module for Robust Speaker Verification of Variable-Duration Utterances
• A Unified Deep Learning Framework for Short-Duration Speaker Verification in Adverse Environments
• Speaker Recognition Based on Deep Learning: An Overview
• Improved End-to-End Dysarthric Speech Recognition via Meta-learning Based Model Re-initialization
• Intra-class variation reduction of speaker representation in disentanglement framework
• Improved Meta-Learning Training for Speaker Verification
• MIRNet: Learning multiple identities representations in overlapped speech
• Learning Metrics from Mean Teacher: A Supervised Learning Method for Improving the Generalization of Speaker Verification System