期刊论文

Wang, ZF

英文

Expert Systems with Applications

0957-4174

2024

244

123004

10.1016/j.eswa.2023.123004

National Key R&D Program of China [2022ZD0117103]; National Natural Science Foundation of China [62077022, 62177022, 62211530433, 62177018, 62277041, 62005092, 62077020, 62173286]; National Teacher Development Collaborative Innovation Experimental Base Construc-tion Research Project of Central China Normal University [CC-NUTEIII 2021-21]; National Nat-ural Science Foundation of Hubei Province, China [2022CFB971]; Shenzhen Science and Technology Program [JCYJ20230807152900001]; Jiangxi Provincial Natural Science Foundation [20232BAB212026]

本校为第一且通讯机构

Transformer models have demonstrated superior performance across various domains, including computer vision, natural language processing, and speech recognition. The success of these models can be attributed to their robust parallel capacity and high computation speed, primarily reliant on the attention layer. In the domain of speaker recognition, state-of-the-art results have been achieved using convolutional neural network (CNN) architectures, particularly with speaker embeddings represented by x-vectors and r-vectors. However, existing CNN-based methods tend to focus on local features while...