发音属性优化建模及其在偏误检测的应用(Speech attributes optimization modeling and application in mispronunciation detection)

no code implementations • CCL 2020 • Minghao Guo, Yanlu Xie

基于GPT-2和互信息的语言单位信息量对韵律特征的影响(Prosodic Effects of Speech Unit’s Information Based on GPT-2 and Mutual Information)

no code implementations • CCL 2022 • Yun Hao, Yanlu Xie, Binghuai Lin, Jinsong Zhang

Text-Aware End-to-end Mispronunciation Detection and Diagnosis

1 code implementation • 15 Jun 2022 • Linkai Peng, Yingming Gao, Binghuai Lin, Dengfeng Ke, Yanlu Xie, Jinsong Zhang

In the field of assessing the pronunciation quality of constrained speech, the given transcriptions can play the role of a teacher.

Contrastive Learning

Speech Enhancement using Separable Polling Attention and Global Layer Normalization followed with PReLU

no code implementations • 6 May 2021 • Dengfeng Ke, Jinsong Zhang, Yanlu Xie, Yanyan Xu, Binghuai Lin

With all these modifications, the size of the PHASEN model is shrunk from 33M parameters to 5M parameters, while the performance on VoiceBank+DEMAND is improved to the CSIG score of 4. 30, the PESQ score of 3. 07 and the COVL score of 3. 73.

Speech Enhancement

A Full Text-Dependent End to End Mispronunciation Detection and Diagnosis with Easy Data Augmentation Techniques

1 code implementation • 17 Apr 2021 • Kaiqi Fu, Jones Lin, Dengfeng Ke, Yanlu Xie, Jinsong Zhang, Binghuai Lin

Recently, end-to-end mispronunciation detection and diagnosis (MD&D) systems has become a popular alternative to greatly simplify the model-building process of conventional hybrid DNN-HMM systems by representing complicated modules with a single deep network architecture.

Data Augmentation

Convolution Forgetting Curve Model for Repeated Learning

no code implementations • 19 Jan 2019 • Yanlu Xie, Yue Chen, Man Li

The status of forgetting (i. e. the output in the working memory model) is regarded as output function or the convolution result of the memory ability and learning material.