The SJTU X-LANCE Lab System for CNSRC 2022

Abstract

This technical report describes the SJTU X-LANCE Lab system for the three tracks in CNSRC 2022. In this challenge, we explored the speaker embedding modeling ability of deep ResNet (Deeper r-vector). All the systems are only trained on the Cnceleb training set and we use the same systems for the three tracks in CNSRC 2022. In this challenge, our system ranks the first place in the fixed track of speaker verification task. Our best single system and fusion system achieve 0.3164 and 0.2975 minDCF respectively. Besides, we submit the result of ResNet221 to the speaker retrieval track and achieve 0.4626 mAP.

Wespeaker: A Research and Production oriented Speaker Embedding Learning Toolkit

“…• ResNet based r-vector and its deeper version, this is the winning system of VoxSRC 2019 [3] and CNSRC 2022 [19].…”

“…The large margin fine-tuning strategy was first proposed in [26] and widely used in speaker verification challenge systems [19,27,28] to further enhance the system's performance. This strategy is performed as an additional fine-tuning stage based on a well-trained speaker verification model.…”

Self-Knowledge Distillation via Feature Enhancement for Speaker Verification

“…For example, [13], [14] introduce a depth-first version of ResNet, significantly boosting the network's depth to an impressive 233 layers. Additionally, [30] reaches a new level by extending ResNet's layer number to 293.…”