The system of speaker verification system shows outstanding performance with the assistance of different types of loss functions with angular margin penalty, which can enforce the intra-class compactness and inter-class discrepancy. However, the power of classification may degrade largely when encountering the cross-domain problems, especially in far-field scenes. Thus, we propose a novel Cross-Domain ArcFace(CD-ArcFace) loss function. By adopting distinct margin penalty in different domain when conducting mix-data fine-tuning, the performance of various speaker verification system can be further improved. This experiment is carried on FFSVC2022. The final score level of our fusion system for the task1 achieves 4.028% and 4.368% EER on the development set and evaluation set.
{"title":"Cross-Domain ArcFace:Learnging Robust Speaker Representation Under the Far-Field Speaker Verification","authors":"Yuke Lin, Xiaoyi Qin, Ming Li","doi":"10.21437/ffsvc.2022-2","DOIUrl":"https://doi.org/10.21437/ffsvc.2022-2","url":null,"abstract":"The system of speaker verification system shows outstanding performance with the assistance of different types of loss functions with angular margin penalty, which can enforce the intra-class compactness and inter-class discrepancy. However, the power of classification may degrade largely when encountering the cross-domain problems, especially in far-field scenes. Thus, we propose a novel Cross-Domain ArcFace(CD-ArcFace) loss function. By adopting distinct margin penalty in different domain when conducting mix-data fine-tuning, the performance of various speaker verification system can be further improved. This experiment is carried on FFSVC2022. The final score level of our fusion system for the task1 achieves 4.028% and 4.368% EER on the development set and evaluation set.","PeriodicalId":282527,"journal":{"name":"The 2022 Far-field Speaker Verification Challenge (FFSVC2022)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114488559","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper presents the development of ZXIC speaker verification system submitted to the task 1 of Interspeech 2022 Far-Field Speaker Verification Challenge (FFSVC2022). Deep neural network based discriminative embeddings, such as x-vectors, have been shown to perform well in speaker verification tasks. In far-field speaker verification system, mismatch between training and testing data and mismatch between enrollment and authentication utterances impact the system performance a lot. To alleviate this mismatch and improve the system performance, in this paper we propose a novel multi-reader domain adaption learning framework based on asymmetric metric learning. In this challenge, we also explore advanced neural network based embedding extractor structures including ECAPA-TDNN and ResNet-SE. A number of experiments on these architectures show that our proposed method is effective and improves the systems performance a lot. The final submitted systems are the fusion of several models. In FFSVC2022, our best system achieves a minimum of the detection cost function (minDCF) of 0.511and an equal error rate (EER) of 4.409 % on the evaluation set.
{"title":"ZXIC Speaker Verification System for FFSVC 2022 Challenge","authors":"Yuan Lei, Zhou Cao, Dehui Kong, Ke Xu","doi":"10.21437/ffsvc.2022-1","DOIUrl":"https://doi.org/10.21437/ffsvc.2022-1","url":null,"abstract":"This paper presents the development of ZXIC speaker verification system submitted to the task 1 of Interspeech 2022 Far-Field Speaker Verification Challenge (FFSVC2022). Deep neural network based discriminative embeddings, such as x-vectors, have been shown to perform well in speaker verification tasks. In far-field speaker verification system, mismatch between training and testing data and mismatch between enrollment and authentication utterances impact the system performance a lot. To alleviate this mismatch and improve the system performance, in this paper we propose a novel multi-reader domain adaption learning framework based on asymmetric metric learning. In this challenge, we also explore advanced neural network based embedding extractor structures including ECAPA-TDNN and ResNet-SE. A number of experiments on these architectures show that our proposed method is effective and improves the systems performance a lot. The final submitted systems are the fusion of several models. In FFSVC2022, our best system achieves a minimum of the detection cost function (minDCF) of 0.511and an equal error rate (EER) of 4.409 % on the evaluation set.","PeriodicalId":282527,"journal":{"name":"The 2022 Far-field Speaker Verification Challenge (FFSVC2022)","volume":"2010 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129127975","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Xiaoyi Qin, Ming Li, Hui Bu, Shrikanth S. Narayanan, Haizhou Li
FFSVC2022 is the second challenge of far-field speaker verification. FFSVC2022 provides the fully-supervised far-field speaker verification to further explore the far-field scenario and proposes semi-supervised far-field speaker verification. In contrast to FFSVC2020, FFSVC2022 focus on the single-channel scenario. In addition, a supplementary set for the FFSVC2020 dataset is released this year. The supplementary set consists of more recording devices and has the same data distribution as the FFSVC2022 evaluation set. This paper summarizes the FFSVC 2022, including tasks description, trial designing details, a baseline system and a summary of challenge results. The challenge results indicate substantial progress made in the field but also present that there are still difficulties with the far-field scenario.
{"title":"The 2022 Far-field Speaker Verification Challenge: Exploring domain mismatch and semi-supervised learning under the far-field scenario","authors":"Xiaoyi Qin, Ming Li, Hui Bu, Shrikanth S. Narayanan, Haizhou Li","doi":"10.21437/ffsvc.2022-3","DOIUrl":"https://doi.org/10.21437/ffsvc.2022-3","url":null,"abstract":"FFSVC2022 is the second challenge of far-field speaker verification. FFSVC2022 provides the fully-supervised far-field speaker verification to further explore the far-field scenario and proposes semi-supervised far-field speaker verification. In contrast to FFSVC2020, FFSVC2022 focus on the single-channel scenario. In addition, a supplementary set for the FFSVC2020 dataset is released this year. The supplementary set consists of more recording devices and has the same data distribution as the FFSVC2022 evaluation set. This paper summarizes the FFSVC 2022, including tasks description, trial designing details, a baseline system and a summary of challenge results. The challenge results indicate substantial progress made in the field but also present that there are still difficulties with the far-field scenario.","PeriodicalId":282527,"journal":{"name":"The 2022 Far-field Speaker Verification Challenge (FFSVC2022)","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123750168","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：