The IDLAB VoxCeleb Speaker Recognition Challenge 2020 System Description

Abstract

In this technical report we describe the IDLAB top-scoring submissions for the VoxCeleb Speaker Recognition Challenge 2020 (VoxSRC-20) in the supervised and unsupervised speaker verification tracks. For the supervised verification tracks we trained 6 state-of-the-art ECAPA-TDNN systems and 4 Resnet34 based systems with architectural variations. On all models we apply a large margin fine-tuning strategy, which enables the training procedure to use higher margin penalties by using longer training utterances. In addition, we use quality-aware score calibration which introduces quality metrics in the calibration system to generate more consistent scores across varying levels of utterance conditions. A fusion of all systems with both enhancements applied led to the first place on the open and closed supervised verification tracks. The unsupervised system is trained through contrastive learning. Subsequent pseudo-label generation by iterative clustering of the training embeddings allows the use of supervised techniques. This procedure led to the winning submission on the unsupervised track, and its performance is closing in on supervised training.

References in corpus (10)

• Scikit-learn: Machine Learning in Python
• A Simple Framework for Contrastive Learning of Visual Representations
• SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition
• Res2Net: A New Multi-scale Backbone Architecture
• VoxCeleb2: Deep Speaker Recognition
• ECAPA-TDNN: Emphasized Channel Attention, Propagation and Aggregation in TDNN Based Speaker Verification
• YouTube-8M: A Large-Scale Video Classification Benchmark
• The IDLAB VoxSRC-20 Submission: Large Margin Fine-Tuning and Quality-Aware Score Calibration in DNN Based Speaker Verification
• Semi-Supervised Contrastive Learning with Generalized Contrastive Loss and Its Application to Speaker Recognition
• Cross-Lingual Speaker Verification with Domain-Balanced Hard Prototype Mining and Language-Dependent Score Normalization

Cited by in corpus (8)

• VoxSRC 2020: The Second VoxCeleb Speaker Recognition Challenge
• Integrating Frequency Translational Invariance in TDNNs and Frequency Positional Information in 2D ResNets to Enhance Speaker Verification
• The IDLAB VoxSRC-20 Submission: Large Margin Fine-Tuning and Quality-Aware Score Calibration in DNN Based Speaker Verification
• The SpeakIn System for VoxCeleb Speaker Recognition Challange 2021
• The JHU submission to VoxSRC-21: Track 3
• The Phonexia VoxCeleb Speaker Recognition Challenge 2021 System Description
• North America Bixby Speaker Diarization System for the VoxCeleb Speaker Recognition Challenge 2021
• Rep Works in Speaker Verification