期刊论文

Zengzhao Chen

英文

Expert Systems with Applications

0957-4174

2023

217

119511

10.1016/j.eswa.2023.119511

The authors thank the editor and anonymous reviewers for their valuable suggestions. This work has been supported by the National Natural Science Foundation of China (Grant No. 62077022 , 61875068 , 62211530433 , 62177018 , 62011530436 , 62277041 , 62005092 , 62077020 ), the National Teacher Development Collaborative Innovation Experimental Base Construction Research Project of Central China Normal University (No. CCNUTEIII 2021-21 ), and the National Key R&D Program of China ( 2021YFC3340802 ), and was supported in part by the National Natural Science Foundation of Hubei Province, China under Grant (No. 2022CFB971 ), the China Unicom Hubei Branch Bilateral Cooperation Research Funds under Grant 2021111002002004 and “Universities Helping Counties” Research Funds of Hubei Province, China under Grant BXLBX0192 .

本校为第一机构

Speaker embeddings have become the most popular feature representation in speaker verification. Improving the robustness of speaker embedding extraction systems is a crucial problem. A multi-scale residual aggregation network (MSRANet), which is a simple but efficient network with triplet input and triplet loss, is proposed in this paper. Two different aggregation strategies are utilized in frame-level feature extractors to capture long-term variations in speaker characteristics. Attention mechanism is employed to filter a large number of parameters in temporal and frequency dimensions, which ...