arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2006.00452eess.AScs.AIcs.CLcs.SD
收藏

基于交叉时延神经网络的说话人识别

Crossed-Time Delay Neural Network for Speaker Recognition

Jilin University(吉林大学);Zhuhai College of Jilin University(吉林大学珠海学院)

Liang Chen, Yanchun Liang, Xiaohu Shi, You Zhou, Chunguo Wu

摘要:

时间延迟神经网络(TDNN)是一种性能良好的基于DNN的说话人识别系统结构。本文提出了一种新的结构交叉时延神经网络(CTDNN),以提高现有TDNN的性能。受卷积神经网络卷积层多滤波器设置的启发,在底层设置多个上下文大小不同的时延单元,构建多层并行网络。建议CTDNN给出了显着的改进,在原来的TDNN的说话人确认和识别任务。在验证实验中,它优于VoxCeleb1数据集,绝对等错误率提高了2.6%。在少炮条件下,CTDNN的识别准确率达到90.4%,比原来的TDNN提高了一倍。我们还将CTDNN与TDNN的另一种新变体FTDNN进行了比较,结果表明,我们的模型在少量射击条件下具有36%的绝对识别精度提高,并且可以在更短的训练时间内更好地处理更大批量的训练,从而更好地利用计算资源。新模型的代码发布于https://github.com/chenllliang/CTDNN

英文摘要:

Time Delay Neural Network (TDNN) is a well-performing structure for DNN-based speaker recognition systems. In this paper we introduce a novel structure Crossed-Time Delay Neural Network (CTDNN) to enhance the performance of current TDNN. Inspired by the multi-filters setting of convolution layer from convolution neural network, we set multiple time delay units each with different context size at the bottom layer and construct a multilayer parallel network. The proposed CTDNN gives significant improvements over original TDNN on both speaker verification and identification tasks. It outperforms in VoxCeleb1 dataset in verification experiment with a 2.6% absolute Equal Error Rate improvement. In few shots condition CTDNN reaches 90.4% identification accuracy, which doubles the identification accuracy of original TDNN. We also compare the proposed CTDNN with another new variant of TDNN, FTDNN, which shows that our model has a 36% absolute identification accuracy improvement under few shots condition and can better handle training of a larger batch in a shorter training time, which better utilize the calculation resources. The code of the new model is released at https://github.com/chenllliang/CTDNN

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑