The SJTU System for Short-duration Speaker Verification Challenge 2021

TL;DR

This paper describes the SJTU system for short-duration speaker verification in the 2021 challenge, utilizing robust embeddings, adaptive scoring, and fine-tuning strategies to improve cross-lingual and phrase-dependent verification accuracy.

Contribution

The paper introduces novel phrase-aware fine-tuning and neural PLDA methods, along with adaptive scoring techniques, to enhance speaker verification performance in challenging scenarios.

Findings

Achieved 0.0473 EER in Task 1 (rank 3)

Achieved 0.0581 EER in Task 2 (rank 8)

Demonstrated effectiveness of phrase-aware strategies

Abstract

This paper presents the SJTU system for both text-dependent and text-independent tasks in short-duration speaker verification (SdSV) challenge 2021. In this challenge, we explored different strong embedding extractors to extract robust speaker embedding. For text-independent task, language-dependent adaptive snorm is explored to improve the system performance under the cross-lingual verification condition. For text-dependent task, we mainly focus on the in-domain fine-tuning strategies based on the model pre-trained on large-scale out-of-domain data. In order to improve the distinction between different speakers uttering the same phrase, we proposed several novel phrase-aware fine-tuning strategies and phrase-aware neural PLDA. With such strategies, the system performance is further improved. Finally, we fused the scores of different systems, and our fusion systems achieved 0.0473 in Task1 (rank 3) and 0.0581 in Task2 (rank 8) on the primary evaluation metric.